论文信息
论文解读
AI 辅助整理- 方法
- 综述AI、IoT传感、机器学习、多模态感知与气候控储技术。
- 发现
- 提出整合数字监测、气候风险评估与可持续害虫治理的六阶段框架。
- 机会点
- 小农户低成本智能储粮、边缘AI与数字孪生、模型跨域迁移是待突破的研究空白。
论文摘要
维持谷物的质量和数量是全球粮食安全的基础。然而,由昆虫和脊椎动物有害生物、储藏霉菌及环境变异性造成的持续采后损失仍在威胁谷物的可用性和安全性。依赖常规检查和化学干预的传统储藏系统正面临气候变化、新发有害生物以及对安全可持续食品需求不断增长的挑战。本文批判性评估了基于人工智能(AI)和物联网(IoT)的传感器、机器学习(ML)、多模态传感、决策支持系统、控温储藏及降低风险的有害生物管理策略的最新进展,重点关注其与预测性管理系统的整合。评估了可持续管理策略的有效性,包括气密储藏、生物农药、硅藻土(DEs)及基于信息化学物质的方法,重点考察其效力、可扩展性和实际可行性,强调基于风险的干预而非全面干预。基于这些进展,本文提出了一个六阶段概念框架,包括储藏条件感知、数据分析、风险预测、决策制定、靶向干预和结果验证。核心理论贡献在于构建了一个整合数字监测、气候风险评估和可持续有害生物管理的概念框架。该框架为储藏管理者识别新风险、选择适当措施和评估管理结果提供了实践指导。主要制约因素包括数据集有限、传感器精度、模型可迁移性、互操作性以及小农户系统中的成本效益。未来优先方向包括推进边缘计算和可解释AI、数字孪生,以及面向资源受限地区的多模态传感和低成本智能储藏解决方案等其他技术。整合数字技术、气候适应和可持续有害生物管理可实现自适应、预测性储粮管理,减少采后损失,提高谷物质量与安全,并增强全球粮食安全。
全文
Frontiers in Sustainable Food SystemsFrontiers in Sustainable Food Systems 01 frontiersin.org
TYPE Review
PUBLISHED 19 November 2026
DOI 10.3389/fsufs.2026.2020632
OPEN ACCESS
EDITED BY
Gurjeet Singh,
Texas A and M University, United States
REVIEWED BY
Miguel Sellitto,
University of the Rio dos Sinos Valley,
Brazil
Ravi Kumar,
Sardar Vallabhbhai Patel University of
Agriculture and Technology, India
CORRESPONDENCE
Hafiz Muhammad Bilal Yousuf
bilal.bashir501@gmail.com
Kurt A. Rosentrater
karosent@iastate.edu
RECEIVED 06 October 2026
REVISED 02 November 2026
ACCEPTED 05 November 2026
PUBLISHED 19 November 2026
CITATION
Yousuf HMB, Yasin M,
Rosentrater KA and Yousuf MU (2026)
Toward smart management of stored
grains: integrating AI, IoT, and
climate-resilient pest management for
global food security.
Front. Sustain. Food Syst. 10:2020632.
doi: 10.3389/fsufs.2026.2020632
COPYRIGHT
© 2026 Yousuf, Yasin, Rosentrater and
Yousuf. This is an open-access article
distributed under the terms of the
Creative Commons Attribution License
(CC BY). The use, distribution or
reproduction in other forums is
permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication
in this journal is cited, in accordance
with accepted academic practice. No
use, distribution or reproduction is
permitted which does not comply with
these terms.
Toward smart management of
stored grains: integrating AI, IoT,
and climate-resilient pest
management for global food
security
Hafiz Muhammad Bilal Yousuf 1, Muhammad Yasin 1,
Kurt A. Rosentrater 2* and Muhammad Usman Yousuf 1
1Department of Entomology, Faculty of Agriculture and Environment, The Islamia University of
Bahawalpur, Bahawalpur, Pakistan, 2Department of Agricultural and Biosystems Engineering, Iowa
State University, Ames, IA, United States
Maintaining the quality and quantity of cereal grains is fundamental to global food
security. However, persistent post-harvest losses caused by insects and verte -
brate pests, storage molds, and environmental variability continue to threaten
grain availability and safety. Conventional storage systems, which require rou -
tine inspection and chemical intervention, are challenged by climate change,
emerging pests, and growing demand for safe and sustainable food. This review
critically evaluates recent advances in artificial intelligence (AI) and Internet of
Things (IoT)-based sensors, machine learning (ML), multimodal sensing, decision-
support systems, climate-controlled storage, and reduced-risk pest management
strategies, with an emphasis on their integration into predictive management
systems. The effectiveness of sustainable management strategies, including her-
metic storage, biopesticides, diatomaceous earths (DEs), and semiochemical-
based approaches, is evaluated with an emphasis on efficacy, scalability, and
practical feasibility, focusing on risk-based rather than blanket interventions.
Based on these advances, the review proposes a six-stage conceptual frame -
work encompassing the sensing of storage conditions, data analysis, risk pre -
diction, decision-making, targeted intervention, and outcome verification. The
central theoretical contribution is the development of a conceptual framework
that integrates digital monitoring, climate risk assessment, and sustainable pest
management. The framework provides practical guidance for storage managers
to identify new risks, choose suitable actions, and assess management outcomes.
Major constraints include limited datasets, sensor accuracy, model transferability,
interoperability, and cost-effectiveness in smallholder systems. Future priorities
include advancing edge and explainable AI, digital twins, and other technologies
such as multimodal sensing and low-cost smart storage solutions for resource-
constrained regions. Integrating digital technologies, climate adaptation, and
sustainable pest management could enable adaptive, predictive stored grain
management, reducing post-harvest losses, improving grain quality and safety,
and strengthening global food security.
KEYWORDS
smart grain storage, predictive analytics, multimodal sensing, AI, climate-resilient
storage, IoT monitoring, sustainable pest control, post-harvest losses
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 02 frontiersin.org
1 Introduction
Post-harvest grain storage plays an essential role in preserving
grain quality and safety (Nath et al., 2024) and ensuring the availability
of stored commodities by bridging seasonal production to a constant
supply (Baidhe et al., 2024). This has become critical for staple grains,
which are stored extensively in order to meet human needs, animal
feed, trade, and national strategic reserves (Nyamai and Anuri, 2025).
Moreover, grain storage is increasingly important for stabilizing
supply chains (Garima et al., 2024 ) and ensuring food availability
during market disruptions and emergencies ( Tushar et al., 2023 ).
However, ensuring grain quality and safety has been a significant chal-
lenge (Osei-Kwarteng et al., 2024), with post-harvest losses posing a
serious threat to food security, particularly in resource-constrained
areas with insufficient storage infrastructure and management capa-
bilities (Shahbazi et al., 2025). Moreover, the urgency of addressing
these challenges is further underscored by the projected increase in
population, which is expected to rise significantly from 7 billion to
9.7 billion by 2050 (Sharma and Sharma, 2025), anticipated to inten-
sify stress on food availability, necessitating an increase in agricultural
production by approximately 70% to meet future demand (Falcon et
al., 2022).
Post-harvest losses can occur in multiple dimensions ( Sokra,
2026), including quantitative losses, manifested through a reduc -
tion in grain weight and volume; qualitative losses, which deterio -
rate the physical and functional properties of grains; and
nutritional losses that are related to the degradation occurring over
an extended period of storage ( Okori et al., 2022 ). Inadequate stor -
age and management practices, particularly in developing coun -
tries, have significantly contributed to the losses, which
compromise food security initiatives ( Bappah and Adejoh, 2024 ).
The effects of these poor practices also lead to adverse economic
consequences, including poor marketability, quality deterioration,
and increased management costs ( Ariong et al., 2023 ). Food safety
hazards are also a major concern, including the proliferation of
fungi and associated mycotoxin production, which can negatively
affect grain quality and safety, pose health risks to humans and
animals, and reduce marketability ( Rhouma et al., 2025 ).
Consequently, post-harvest losses represent not only a loss of food
but also a waste of resources such as land, water, and labor, since
these resources constitute important inputs to grain production
(Shahbazi et al., 2025 ). Moreover, the physical grain losses are esti -
mated to be between 10 and 20% (and up to 25–30% in some tra -
ditional systems) in developing countries ( Okori et al., 2022 ;
Y ousuf et al., 2026b).
The causes of stored grain losses are multifaceted and often
interrelated ( Y ang et al., 2026). The most devastating stored grain
insect pests are predominantly found in the orders Coleoptera,
Lepidoptera, and Psocodea, which include beetles, moths, and
psocids ( Demis and Y enewa, 2022 ). Coleopteran pests, such as
Sitophilus spp., Rhyzopertha dominica (Fabricius; Coleoptera:
Bostrichidae), Trogoderma granarium Everts (Coleoptera:
Dermestidae), and Tribolium spp., are responsible for inflicting
direct feeding damage, weight loss, contamination, and degrada -
tion of grain quality ( Chandaragi et al., 2023 ; Khan and Khan,
2023). Similarly, the Lepidopteran pests that commonly include
Sitotroga cerealella (Olivier; Lepidoptera: Gelechiidae) and Plodia
interpunctella (Hübner; Lepidoptera: Pyralidae) also cause signifi -
cant losses to stored products ( Ahmad et al., 2021 ; Chowdhury et
al., 2023 ). Psocids, mainly Liposcelis spp., are commonly associated
with humid or high-moisture storage conditions and can also
cause serious losses ( Y ousuf et al., 2025). Temperature and mois -
ture fluctuations, including moisture migration and condensation,
can accelerate grain respiration, localized heating, mold growth,
and subsequent quality degradation ( Liu C. et al., 2025 ). Moreover,
rodent attacks result in substantial physical losses and contamina -
tion of the grain ( Prasad et al., 2025 ), while molds produce myco -
toxins ( Okori et al., 2022 ). Repeated application of chemical
pesticides and fumigants increases the risk of residues, occupa -
tional and environmental hazards ( Guru et al., 2022 ), and target-
insect resistance, reducing the effectiveness of regular management
practices ( Y ousuf et al., 2026b). These challenges are further exac -
erbated by infrastructure and managerial inadequacies ( Ariong et
al., 2023 ), particularly in smallholder storage systems, which face
limited access to appropriate facilities, monitoring technologies,
and technical expertise ( Nakoma-Ngoma et al., 2025b ).
Conventional grain storage methods conventional grain storage
methods, which are based on regular inspections and reactive inter-
ventions (Chen, 2025), are increasingly challenged by the changing
climate, extreme weather events, altered pest dynamics, extended
storage periods, and the need for safe and sustainable food sources
(Latha, 2026). Increasing temperatures and moisture levels can expe-
dite pest development and affect storage conditions, whereas
extreme weather can interfere with transportation and storage oper-
ations (Gerken and Morrison III, 2022 ). In parallel, growing con -
cerns about injudicious use of pesticides, the harmful effects of their
residues, environmental impact, and regulatory restrictions under -
score the need for more targeted, environmentally friendly, and less
hazardous interventions ( Guru et al., 2022 ). Therefore, detecting
deterioration only after damage has occurred is no longer compati -
ble with the need for preventive and cost-effective grain manage -
ment (Y ang et al., 2026). A bibliometric analysis of post-harvest
grain activities highlights the relationship between digital technolo -
gies, sustainability, and efficiency within the broader context of
research trends on digital transformation ( Schmidt et al., 2024 ).
Similarly, another case study demonstrated opportunities to trans -
late digital innovations into practical post-harvest applications, par-
ticularly improving grain storage operations and supporting better
management decisions (Schmidt et al., 2023).
These complications have sparked interest in smart and sustain -
able management systems that incorporate real-time monitoring,
predictive analytics, climate adaptation, and risk-reduction strate -
gies ( Alyammahi et al., 2025 ; Latha, 2026 ). Internet of Things
(IoT)-enabled sensors, artificial intelligence (AI), and machine-
learning (ML) can be used to monitor temperature, relative humid -
ity, grain moisture, carbon dioxide, pest incidence, and other
factors affecting grain condition ( Fuentes-Peñailillo et al., 2024 ;
Luo et al., 2025 ). Such data can enable early detection, risk assess -
ment, and prediction of pest outbreaks and mold growth
(Alyammahi et al., 2025 ), as well as grain deterioration and storage
failures (Sharma and Shivandu, 2024 ). Moreover, sustainable inte -
grated pest management (IPM) practices such as hermetic storage,
diatomaceous earths (DEs), botanicals, biopesticides, biological
control, and semiochemical strategies can reduce reliance on tradi -
tional chemical treatments ( Hembram et al., 2025 ; Turabi et al.,
2026). Climate-resilient storage technologies such as enhanced
insulation, aeration, low-energy cooling, moisture management,
and renewable energy are promising innovations that can improve
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 03 frontiersin.org
the adaptability of storage systems to a changing environment
(Y ousuf et al., 2026b).
This progression is also compatible with international initia -
tives to reduce food losses, improve food safety ( Okori et al.,
2022), enhance climate adaptation, and promote sustainable food
systems ( Turabi et al., 2026 ). The Food and Agriculture
Organization (FAO) emphasizes reducing food loss and waste to
strengthen food security and promote sustainable agrifood sys -
tems ( FAO, 2026 ), while the United Nations Sustainable
Development Goals (SDGs) highlight food security, sustainable
agriculture, responsible consumption, and climate resilience
(United Nations, 2026 ). Therefore, it is a critical contribution to
these international priorities to improve the performance, safety,
and resilience of grain storage systems ( Baidhe et al., 2024 ; Garima
et al., 2024 ).
Although significant advances have been made in understand -
ing stored-grain pests and emerging management technologies,
the existing literature remains highly fragmented across these
research areas. Studies on digital technologies have explored the
potential of sensors, IoT, and AI to monitor and predict grain
quality and storage conditions ( Lutz and Coradi, 2022 ). Similarly,
several studies have used mathematical modeling techniques to
understand and mitigate risk in stored grain systems
(Anukiruthika and Jayas, 2024 ), while others have focused on
opportunities for digital innovations in post-harvest systems
(Schmidt et al., 2023 ). Despite these developments, there is a lack
of integration between biological knowledge, monitoring, predic -
tive risk assessment, and sustainable control measures. In particu -
lar, there is a need to integrate the identification of potential
storage problems with the selection and implementation of appro -
priate control measures and the evaluation of their effectiveness.
This is an important limitation because individual monitoring and
prediction techniques may not provide timely and specific man -
agement decisions across different storage and climatic condi -
tions. To bridge this knowledge gap, the current review highlights
the critical linkages among these components. It presents a con -
ceptual six-stage framework that integrates sensing, analysis, pre -
diction, decision-making, intervention, and verification for smart,
predictive, and climate-resilient stored-grain management. The
proposed framework provides a basis for linking early risk iden -
tification with focused, sustainable interventions to improve grain
quality and safety, reduce post-harvest losses, and ensure global
food security. This article is a conceptual review rather than an
empirical investigation and does not present original experimen -
tal findings. The arguments are developed through a critical syn -
thesis of published literature, focusing on the relationships,
limitations, and practical implications of existing research.
The review begins by examining storage systems, grain dete -
rioration, and pest-related challenges to establish the basis for
effective management. It then addresses climate-related risks and
adaptation strategies, followed by digital technologies that support
improved monitoring and decision-making. Sustainable and
reduced-risk protection approaches are evaluated to identify their
practical relevance in stored-grain management. These discus -
sions establish the foundation for the proposed integrated concep -
tual framework, linking the essential components of predictive
and climate-resilient grain management. Implementation chal -
lenges and future research priorities are subsequently discussed to
identify opportunities for broader adoption.
2 Storage systems and drivers of
quality loss
2.1 Storage systems and management
practices
Grain storage systems are characterized by a wide range of struc-
tural variations, from traditional household-level systems to advanced
and highly digitized structures, which vary in their storage capacity,
control mechanisms, monitoring levels, and management intensity
(Wu et al., 2023; Nickhil et al., 2025). These systems can be further
categorized into conventional, improved and reduced-risk, climate-
resilient, and predictive grain storage systems (Okparavero et al., 2024;
Jain et al., 2026). Conventional systems generally offer limited control
over temperature, moisture, and pest invasion, while improved and
reduced-risk systems incorporate enhanced sealing, atmospheric
modification, or other approaches to minimize deterioration and reli-
ance on chemicals (Chen, 2025; Madhu, 2025). Climate-resilient sys-
tems are capable of regulating temperature and moisture levels
through enhanced aeration, insulation, cooling mechanisms, and
moisture control (Iqbal et al., 2025). Innovative smart-storage systems,
empowered with IoT-enabled sensors and AI/ML-driven surveillance
capabilities, ensure continuous data acquisition, enable early warning
detection, and facilitate risk-based management (Lutz and Coradi,
2022; Luo et al., 2026; Table 1).
Storage performance depends not only on structural properties
but also on grain condition at intake, initial moisture content (MC),
cleanliness, aeration, and storage period (Ziegler et al., 2021; Jain et
al., 2026). These factors determine the stability of the storage environ-
ment and the rate and extent of quality deterioration (Okparavero et
al., 2024). Accordingly, storage technologies should be viewed as com-
ponents of an integrated management system rather than as isolated
interventions.
2.2 Drivers and consequences of grain
deterioration
Stored grain deterioration is primarily caused by the combination
of biological, physical, chemical, and environmental factors, which
can result in adverse economic and food-safety implications (Ziegler
et al., 2021; Y ang et al., 2026). Biotic agents, including insect pests,
mold, and rodents, are among the most common causes of grain
weight loss, degradation, and contamination during storage (Okori et
al., 2022; Chandaragi et al., 2023 ). Insects infest grains by feeding
directly and damaging kernels; molds modify grain composition and
produce mycotoxins under suitable environmental conditions, while
rodents consume, contaminate, and damage grain structures (Y aseen
et al., 2019; Ahmad et al., 2021). The biological activity of insects and
the grain itself contributes to increases in heat and moisture, thereby
promoting degradation (Stathas et al., 2023) as shown in Figure 1.
Temperature and moisture are the major factors that determine
grain stability (Tushar et al., 2023). The hygroscopic nature of grain
often creates a dynamic equilibrium between moisture in the grain
and the air, and thermal variation within bulk storage results in mois-
ture transfer and condensation (Yin et al., 2025a). These hotspots pro-
mote grain respiration and create conditions for pest proliferation and
mold growth (Baidhe et al., 2024). Inadequate aeration can promote
these effects, while proper temperature and moisture control can
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 04 frontiersin.org
TABLE 1 Comparison of major stored grain management systems.
Category Storage
systems
Application Environmental
control
Deterioration risks Monitoring
capability
Advantages Limitations Literature
cited
Traditional and
conventional systems
Bags, bins, pits,
warehouses, silos,
grain elevators
Household, smallholder,
farm, community, and
commercial
Low to moderate; higher in
engineered facilities
Insects, moisture
fluctuations, condensation,
fungi, rodents, hotspots
Mainly manual or
periodic; automated
monitoring possible in
larger facilities
Widely available;
flexible; established
handling infrastructure
Limited early detection;
environmental
variability; moisture
and pest exposure;
higher losses when
management is
inadequate
Kumar and Kalita
(2017); Bhardwaj and
Sharma (2020);
Olorunfemi and
Kayode (2021); Singh
and Jayas (2024)
Improved and
reduced-risk systems
Hermetic bags,
hermetic cocoons,
sealed bins,
modified/
controlled-
atmosphere storage
Smallholder,
cooperative, farm, and
specialized/commercial
Moderate to high; passive or
actively modified
atmosphere
Insects, residual moisture,
fungal deterioration if
grain is poorly conditioned
Manual or sensor-
compatible; gas
monitoring in
specialized systems
Reduced chemical
dependence; improved
pest protection;
adaptable to different
scales
Requires adequate
grain drying and
sealing; effectiveness
depends on integrity,
moisture management,
and technical
requirements
Navarro (2012);
Baributsa et al.
(2017); Baributsa and
Concepcion Ignacio
(2020); Y ousuf et al.
(2025)
Climate-resilient
systems
Improved aeration,
insulation, low-
energy cooling,
solar-assisted
cooling, moisture-
management
systems
Farm, community,
commercial, and
industrial
Moderate to high; active
temperature/moisture
management
Grain heating, moisture
migration, condensation,
insects, fungal growth
Increasingly sensor-
enabled
Helps stabilize
temperature and
moisture; improves
resilience to climatic
variability
Energy, infrastructure,
maintenance, and
climate-specific design
requirements
Li et al. (2020);
Tripathy et al. (2025);
Zhang L. et al. (2026);
Smart and predictive
systems
IoT-enabled storage,
AI/ML-supported
monitoring,
predictive storage
platforms, digital-
twin-enabled
systems
Commercial and
emerging smallholder
applications
High/adaptive; continuous
environmental and analytical
control
Pest infestation, hotspots,
moisture changes, fungi,
grain-quality deterioration
Continuous/real-time
with analytical and
predictive capability
Remote monitoring,
automated alerts, early
warning, risk-based
management, adaptive
decision support
Data and connectivity
requirements; sensor
calibration and
maintenance; model
validation and
interoperability
challenges; technical
expertise and cost
Jain et al. (2026); Lutz
and Coradi (2022);
Lydia et al. (2022);
Fuentes-Peñailillo et
al. (2024); Y ousuf et
al. (2026b)
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 05 frontiersin.org
maintain grain quality ( Kulazhanov et al., 2024 ; Chelladurai et
al., 2025).
Chemical and nutritional deterioration can occur even without
visual evidence of infestation. Extended storage periods can lead to
lipid oxidation, protein and carbohydrate degradation, the formation
of off-flavors, and impaired germination capacity ( Ali et al., 2022 ;
Zhang Y . et al., 2025). Therefore, grain quality and nutritional value
decline, while technological properties such as milling, flour, dough,
and baking qualities are also affected, making it less suitable for food
or seed purposes (Atanda et al., 2016; Khalid et al., 2023).
The consequences of deteriorating the quality of agricultural com-
modities are not limited to physical losses. A decline in quality reduces
market value, increases cleaning and treatment costs, shortens storage life,
and can lead to rejections or price deductions (Odjo et al., 2022; Stathas
et al., 2023). Food-safety risks, such as fungal contamination, mycotoxin
accumulation, insect fragments, rodent activity, and pesticide residues,
are equally concerning and can render grain unsuitable for human or
animal consumption (Gurikar et al., 2023; Prasad et al., 2025). Therefore,
effective storage management should focus on maintaining the quantity
and quality of stored grain to ensure that they remain safe and suitable for
human and animal consumption throughout the storage period.
2.3 Limitations of conventional monitoring
Conventional grain storage practices mostly rely on manual sam-
pling and periodic inspections, as well as decision-making based on
predetermined schedules or visual assessment of the observed deterio-
ration (Chen, 2025). These methods usually have low temporal and
spatial resolutions and are unable to detect early and localized changes
in grain condition. As a result, management interventions can only be
initiated after significant degradation has already occurred
(Yunshandan et al., 2025).
The limitations of periodic monitoring tend to be more evident
under different storage conditions or during extended storage.
Continuous monitoring of different parameters such as MC, tempera-
ture, and gas concentrations can provide greater situational awareness
and support more timely prediction and intervention (Labrot-Rhodes
et al., 2025). These limitations can serve as a foundation for recogniz-
ing the importance of continuous monitoring as well as a risk-based
assessment, which in turn serves as the basis for the detailed consid-
eration of stored grain pests and their management.
3 Stored grain pests, management, and
risk prediction
3.1 Major pests and environmental drivers
Stored grain pests are major biological constraints to grain quality
and stability during storage. Their infestation is influenced by feeding
behavior, grain quality, and storage conditions (Demis and Y enewa,
2022; Stathas et al., 2023). Stored grain insect pests are broadly catego-
rized as primary pests, which attack intact kernels, and secondary pests,
which infest damaged or processed grain (Agrafioti et al., 2023; Hassan
et al., 2023; Ponce et al., 2024). Other biotic hazards, including storage
mites, rodents, and toxigenic fungi, are also responsible for the degrada-
tion of stored grains and pose threats to food safety (Y ousuf et al., 2026b;
Table 2). The development of these pests depends heavily on tempera-
ture, relative humidity, grain MC, storage time, and grain condition.
FIGURE 1
Major drivers of stored grain losses.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 06 frontiersin.org
Favorable conditions promote reproduction rates and survival, increas-
ing the likelihood of kernel damage and subsequent colonization
(Chaudhary et al., 2025). This can increase insect respiration and heat
production, creating favorable environments for infestation (Stathas et
al., 2023; Baidhe et al., 2024). Moreover, climate change could influence
pest development and infestation processes, further emphasizing the
need for adaptive surveillance and management strategies (Y adav et
al., 2024).
3.2 Conventional pest management and
insecticide resistance
Conventional pest management practices mainly rely on the
use of fumigation, residual insecticides, sanitation, exclusion,
and physical removal of the pests ( Guru et al., 2022 ). Although
these strategies can significantly reduce pest infestations, there
are growing concerns about the negative impacts associated with
intensive chemical use, such as development of resistant strains,
pesticide residues, occupational exposure, as well as adverse
environmental impacts, which have led to limitations in their use
(Murugesan et al., 2024 ). Resistance may develop through altera -
tions in the target site, increased metabolic breakdown, reduced
penetration, or behavioral changes in insects. Resistance man -
agement strategies include monitoring susceptibility, proper
application, rotation of tactics, and avoidance of unnecessary
treatment ( Al Naggar et al., 2025 ). These limitations highlight
the need for more focused, risk-based strategies for insect pest
management.
TABLE 2 Major stored grain pests and management strategies.
Grain
type
Major insect
pests
Other biotic
stressors
Deterioration-risk
conditions
Management and
predictive
opportunities
Literature cited
Wheat R. dominica, T.
granarium, Sitophilus
spp., Tribolium spp.,
Oryzaephilus
surinamensis
(Linnaeus), Ephestia
spp., Liposcelis spp.
Storage mites; rodents;
fungi; DON and
associated mycotoxins
~13–13.5% moisture;
RH < 65–70% desirable; higher
temperature, moisture and
prolonged storage increase
deterioration risk
Hermetic storage, fumigation,
DE, botanicals, IPM; acoustic,
carbon dioxide (CO2) and
volatile organic compounds
(VOCs) sensing; imaging; ML
models
Okori et al. (2022); Stathas et
al. (2023); Diab et al. (2025);
Mutalov et al. (2025); Y ousuf et
al. (2026b)
Rice Sitophilus oryzae
(Linnaeus), S. cerealella,
Tribolium spp., Corcyra
cephalonica (Stainton),
P . interpunctella,
Liposcelis spp.
Storage mites; rodents;
fungi; aflatoxins and
OTA
~13% moisture for long-term
milled-rice storage; ≤14% for
shorter storage; RH < 65–70%
desirable; high moisture and
temperature increase insect/
fungal risk
Hermetic storage, controlled
atmospheres, sanitation, IPM;
CO2/acoustic sensing;
imaging; environmental
monitoring
Navarro (2012); Manandhar et
al. (2018); Lutz et al. (2022);
Khan (2024); Athanassiou et al.
(2025b); Mutalov et al. (2025);
Li W . et al. (2025); Jain et al.
(2026)
Maize Sitophilus zeamais
Motschulsky,
Prostephanus truncatus
(Horn), Tribolium spp.,
S. cerealella, Ephestia
spp., Liposcelis spp.
Storage mites; rodents;
fungi; aflatoxins and
fumonisins
~13–13.5% moisture;
RH < 65–70% desirable;
elevated temperature, moisture
and kernel damage increase
insect and fungal risk
Hermetic storage,
biopesticides, sanitation,
IPM; acoustic/CO2 sensing;
imaging; predictive modeling
Stathers et al. (2020); Khan
(2024); Athanassiou et al.
(2025b); Madhukar et al.
(2025); Shimbwadala et al.
(2025); Abdel Halim et al.
(2026); Jain et al. (2026);
Barley R. dominica, Sitophilus
spp., Tribolium spp.,
Oryzaephilus spp.,
Ephestia spp., Liposcelis
spp.
Storage mites; rodents;
fungi; DON and other
mycotoxins
~12–13% moisture generally
targeted; RH < 65–70%
desirable; higher temperature
and moisture favor fungal and
insect activity
Aeration, hermetic storage,
IPM; temperature/moisture
sensing; predictive modeling
Khan (2024); Athanassiou et al.
(2025b); Zhang X. et al. (2026)
Sorghum Sitophilus spp., T.
granarium, Ephestia
cautella (Walker),
Liposcelis spp.,
Tribolium spp., C.
cephalonica
Storage mites; rodents;
fungi; aflatoxins and
other mycotoxins
~12.5–13.5% moisture for
long-term storage; RH < 65–
70% desirable; increasing
temperature and moisture
enhance deterioration risk
Hermetic storage, heat
treatment, biological control,
IPM; acoustic/image-based
detection
Kumar and Kalita (2017);
Manandhar et al. (2018); Abay
et al. (2025); Madhukar et al.
(2025)
Millet S. cerealella, Tribolium
spp., Ephestia spp., C.
cephalonica, Liposcelis
spp., Oryzaephilus spp.
Storage mites; rodents;
fungi; mycotoxins
~13–16% moisture depending
on grain type and storage
duration; RH < 65–70%
desirable; higher moisture and
temperature increase
deterioration risk
Hermetic storage, sanitation,
biological control, IPM;
low-cost sensing; early-
warning systems
Kumar and Kalita (2017); Khan
(2024); Athanassiou et al.
(2025b); Mutalov et al. (2025);
Hou et al. (2026)
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 07 frontiersin.org
3.3 Sustainable IPM
Sustainable IPM employs integrated strategies that reduce pest infes-
tation risks while minimizing the hazards of chemical contamination
(Zhou et al., 2024). Hermetic and controlled-atmosphere techniques
prevent insect infestations, while DEs, botanicals, biopesticides, biologi-
cal control, and various other physical approaches provide alternative
pest control measures (Abdel Halim et al., 2026). Effective IPM imple-
mentation necessitates the combined use of sanitation, removal of
infested and contaminated grains, and other structural and mechanical
interventions (Bingham and Hagstrum, 2023). Since the effectiveness of
a given strategy mostly depends on the pest complexes, grain types, and
environmental conditions, the application of IPM practices should be
optimized through risk assessment to ensure the best control measures
for a particular situation and target pests (Scheff and Phillips, 2022).
3.4 Early detection and digital pest
surveillance
Traditional inspection methods may be limited in their ability to
identify early-stage or internal infestations, particularly when the pest
population exists in a specific zone of the storage facility (Keerthana et al.,
2026). Advanced technologies, such as acoustic detection, carbon dioxide
(CO2) monitoring, hyperspectral and near-infrared imaging, volatile
organic compounds (VOCs), and biosensors, offer promising alternatives
for non-invasive, near-real-time monitoring of stored products (Prempeh
et al., 2025; Haque et al., 2026). Combined with digital monitoring, such
methods have the potential to enhance detection capabilities and data
resolution, and provide the opportunity to quantify the risk of pest
invasion.
3.5 Predictive pest management and
emerging molecular approaches
Predictive pest management practices are focused on the iden -
tification of infestation risk before the population reaches an
economically damaging threshold ( Ibrahim et al., 2022 ; Mittal et
al., 2024 ). Degree-day and population models can use tempera -
ture, humidity, grain MC, storage period, historical information,
and current observations to predict pest development and the
likelihood of outbreaks ( Day et al., 2024 ; Abduljaleel et al., 2025 ).
The application of weather data along with ML to surveillance and
response activities will enhance climate-informed forecasting and
facilitate the rapid implementation of interventions ( Paria and
Das, 2026 ).
Emerging molecular techniques, including RNA interference
(RNAi) and clustered regularly interspaced short palindromic
repeats (CRISPR), provide species-specific control ( Faber et al.,
2025). However, their role in the case of stored grain pests needs
to be evaluated experimentally, and suitable delivery systems
should also be developed for their successful implementation
(Y adav et al., 2025). Altogether, environmental monitoring, early
detection, and forecasting offer the possibility of shifting from
reactive to more proactive interventions in stored grain pest man -
agement ( Figure 2 ).
4 Climate change and storage
vulnerabilities
Changing climate patterns pose a threat to the stability and
security of stored grain around the globe by modifying tempera -
tures and humidity, encouraging pest proliferation, promoting
fungal growth, and increasing the likelihood of extreme weather
events ( Y an et al., 2026). These changes can result in increased
degradation of stored grain and complicate the storage practices
that were originally established in less varied climates ( Y ousuf et
al., 2026b ). Thus, understanding interactions between
climate-related changes and storage conditions is important in
order to design robust and adaptable grain-management
systems.
FIGURE 2
Stored grain pest risk management continuum.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 08 frontiersin.org
4.1 Climate-driven pest dynamics
Variations in temperature regimes and seasonal patterns may
affect the developmental and reproductive processes, abundance, sur-
vival, and distribution of stored grain pests (Gerken and Morrison III,
2022). An increase in temperatures within suitable thermal ranges can
accelerate pest development and reduce generation times, whereas
changes in seasonal regimes may influence the timing and persistence
of infestations (Guru et al., 2022). High grain temperature could lead
to increased respiration and create localized heating, resulting in a
microclimate conducive to the activity of pests and further secondary
deterioration if aeration is inadequate (Zhao et al., 2024; Y ang et al.,
2026). Hence, climate-induced changes in pest populations may dis-
rupt the timing and efficacy of conventional management practices,
underscoring the need for adaptive monitoring and risk-based
interventions.
4.2 Moisture variability, fungal growth, and
mycotoxin risks
Variations in relative humidity lead to changes in the equilibrium
MC of grains, causing increased moisture transfer and condensation
within the grain structure (Yin et al., 2025b). In addition, fluctuations
in MC (particularly during wetting and drying cycles) negatively
affect grain quality, while favorable MC and temperature promote
fungal proliferation and toxin production (Lutz et al., 2022; Saeed et
al., 2023; Zheng et al., 2024). The impact of these hazards depends
primarily on the moisture level and temperature of stored grain,
duration of storage, grain condition, and fungal species. The major
mycotoxins associated with storage commodities are aflatoxins,
deoxynivalenol (DON), fumonisins, and zearalenone, which pose
significant health and economic risks to grain producers (Gurikar et
al., 2023).
4.3 Extreme weather vulnerabilities and
climate-resilient storage
Flooding, heatwaves, droughts, wet harvests, and power out -
ages affect drying, transportation, storage infrastructure, and
environmental control ( Narayan et al., 2024 ; Onyeaka, 2026 ).
High grain MC accelerates deterioration, whereas heatwaves
increase cooling requirements and reduce the efficacy of passive
cooling ( Rosentrater, 2022 ; Potoroko et al., 2024 ). Power disrup -
tions also affect aeration, cooling, and automated monitoring sys -
tems ( Lutz et al., 2022 ; Zhang X. et al., 2026 ). Therefore, storage
with enhanced resilience to climate variability requires improved
facilities as well as contingency measures against environmental
and infrastructural disruptions. Adaptation in grain storage
includes technologies and practices that regulate moisture and
temperature while minimizing reliance on energy- or chemical-
intensive strategies ( Labrot-Rhodes et al., 2025 ).
Hermetic storage changes the gas composition within the stor -
age facility to prevent or suppress insect development, whereas insu-
lation, aeration, moisture regulation, and solar-assisted cooling may
help manage temperature and humidity. These adaptive technologies
depend on the grain type, local climate, storage scale, infrastructure,
and managerial capacity, and there is a need to consider these
factors when choosing appropriate measures (Okori et al., 2022). In
terms of strategies, the focus in resource-constrained areas should
be on cost-effective practices, such as proper drying, improved seal -
ing, hermetic storage, aeration, and affordable monitoring, whereas
commercial storage facilities could prioritize automated cooling,
sensors, and advanced climate control mechanisms (Yunshandan et
al., 2025). The practical implementations of these interventions in
the field have already demonstrated their efficacy in reducing grain
temperature. For example, in Hangzhou, China, the use of a radia -
tive-cooling membrane in grain storage warehouses led to a reduc -
tion in the headspace temperature by 9.8 °C and in the grain
temperature by 4 °C, demonstrating the effectiveness of passive
interventions in reducing the risk of heat stress in stored grain ( Xu
et al., 2023).
4.4 Climate-informed risk prediction
Climate-adaptive strategies depend heavily on anticipating rather
than reacting to storage risks ( Muccione et al., 2024 ). The use of
weather forecasts, real-time monitoring of the conditions within the
storage environment, and knowledge of pest and fungal biology can
predict the conditions conducive to insect development, fungal pro-
liferation, imbalances in moisture, and grain quality degradation
(Bukhari et al., 2025). The combination of such data using statistical
or ML methods can help to evaluate risks and take timely measures
(Pinto et al., 2025). Therefore, climate-informed forecasts represent a
critical link between conventional storage management practices and
digital technologies (Figure 3).
5 Digital transformation of stored grain
management
Digital technologies are transforming stored grain management
practices by incorporating sensing, IoT, AI, and data analytics into
storage facilities (Schmidt et al., 2025; Y ang et al., 2026). These inno-
vative technologies have the potential to continuously monitor the
conditions of stored grain, detect and predict deterioration risks early,
and support timely decision-making across diverse storage and cli-
matic conditions (Amiri, 2026; Latha, 2026). Recent research high-
lights the role of AI and digital technologies in monitoring storage
conditions and supporting predictive grain management
(Anukiruthika and Jayas, 2025). Similarly, stored-grain pest monitor-
ing and decision-support systems provide a basis for linking risk
assessment with targeted management interventions (Athanassiou et
al., 2025a). These constructs were selected and organized based on
their respective roles in identifying storage risks, assessing potential
threats, guiding management interventions, and evaluating their
outcomes.
5.1 From periodic inspection to continuous
sensing
Traditional grain storage management approaches are typically
associated with conventional sampling techniques and regular inspec-
tions that may not always be effective in detecting emerging or
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 09 frontiersin.org
localized problems (Kulazhanov et al., 2024). Digital storage solutions,
on the other hand, are capable of providing constant or near-real-time
information about conditions within the storage facility as well as the
biological processes, allowing timely detection of any damage. In addi-
tion, continuous and prolonged data acquisition facilitates the devel-
opment of predictive models that can forecast the need for responsive
and adaptive management (Labrot-Rhodes et al., 2025).
5.2 Sensor technologies and IoT-enabled
monitoring
Modern sensors can monitor and record critical parameters
related to storage conditions, including temperature, relative
humidity, grain MC, concentrations of oxygen (O 2) and CO 2,
acoustic activity, VOCs, and visual/spectral features ( Labrot-
Rhodes et al., 2025 ; Zhang L. et al., 2026 ). Physical sensors mea -
sure storage environment parameters, while gas, acoustic, VOCs,
and imaging or spectral data provide indirect or direct evidence
of insect activity, fungal development, or changes in grain charac -
teristics ( Gong et al., 2026 ). Combining these indicators can pro -
vide a better understanding of the current storage conditions than
reliance on a single variable. In addition, these data can be utilized
to build an automated detection system and predict possible
changes. However, the performance of CO 2 and VOCs monitoring
systems during localized spoilage depends on sensor location and
gas flow dynamics within the grain mass. Spatial variability in gas
concentrations, diffusion, or airflow patterns can delay or reduce
the identification of deterioration, especially when sensors are
located far from the affected areas ( Nunes et al., 2023 ).
IoT frameworks typically use wireless communication net -
works to link sensors to cloud- or edge-based platforms to store,
process, and visualize data ( Viviane et al., 2023 ; Olatinwo et al.,
2025). IoT-based monitoring systems have been integrated to
enable the collection of real-time storage data and the use of pre -
dictive models for grain quality assessment ( Coradi et al., 2022 ).
These systems can allow for automated alerts, remote access, and
long-term data management that reduce the need for routine
manual inspections of grain storage facilities in response to identi -
fied irregularities ( Mchinzi and Banda, 2025 ). Edge computing
can be used for local data management in the absence of network
connectivity, while cloud platforms are helpful in storing and ana -
lyzing information. Their effectiveness is determined by the reli -
ability of sensors, energy sources, connections, maintenance, and
cost-effectiveness ( Kuchuk and Malokhvii, 2024 ; Yu et al., 2025 ).
5.3 AI and ML for detection and prediction
AI algorithms and ML techniques can be used to analyze the
information obtained from different types of sensors, images, acous-
tics, and spectra to identify patterns linked to insect infestation, grain
damage, fungal contamination, and quality deterioration (Murmu et
al., 2026). Although computer-based classification offers better effi-
ciency and is faster than the conventional manual approach, the accu-
racy of such techniques relies on the quality and quantity of the
datasets used (Aviara et al., 2022; Luo et al., 2026).
In addition to monitoring and detection, predictive models can esti-
mate the likelihood and timing of future pest infestation, spoilage,
FIGURE 3
Climate change impacts on stored grain systems.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 10 frontiersin.org
moisture fluctuations, or mycotoxin production. Based on the available
information, predictive models can estimate the possibility or timing of
future pest infestation, spoilage, moisture fluctuations, or mycotoxin
production (Rodrigues et al., 2024; Latha, 2026). By combining informa-
tion about environmental factors, historical data, and climatic influences,
it is possible to obtain more accurate estimates of these parameters and
take timely measures to prevent the identified risks and to shift post-
harvest grain management from a reactive to a proactive approach
(Gerken and Morrison III, 2022; Y ang et al., 2026).
5.4 AI-based decision-support systems
The predictive outcomes can be made practically applicable by
developing them into management recommendations. Decision-support
systems based on AI models can integrate predicted risk with the current
storage conditions and thresholds to recommend specific control mea-
sures such as aeration, cooling, drying, or targeted pest control treat-
ments (Sahu et al., 2025 ; Luo et al., 2026 ). This enables risk-based
management, where the intensity and timing of interventions are based
on the predicted risk instead of following a fixed schedule (Garima et al.,
2024). These functions can be used to support risk-based decision-mak-
ing in the digital management pathway (Figure 4).
5.5 Computer vision tools
Computer vision approaches, such as red, green, and blue (RGB),
hyperspectral, and near-infrared (NIR) image sensing, can be used in
non-destructive quality assessment of grain appearance, contamination,
insect damage, and other quality indicators (Liu et al., 2022; Ageh et al.,
2025). In addition, the use of image-based systems can be implemented
in automated grading and classification of grains, which is promising,
particularly with the development of smartphone applications. However,
the widespread application of such technologies is still limited and
depends on their ability to provide consistent results across diverse grain
types under varying lighting conditions, image resolutions, and environ-
mental parameters (Kuntke et al., 2024; Siddika et al., 2025).
5.6 Digital twins and predictive simulation
Digital twins can create a virtual representation of grain storage sys-
tems by combining their structure with real-time and historical data
(Peladarinos et al., 2023; Zhang R. et al., 2025). These virtual versions can
replicate temperature and moisture changes, pest invasions, and climatic
shifts, thus allowing management options to be assessed before their on-
site application (Dai et al., 2024). By linking sensor data with virtual
copies, digital twins can identify emerging abnormal conditions, explore
alternative aeration or cooling configurations, and estimate their effect
on grain quality without disturbing the stored grain. They can also evalu-
ate different weather scenarios and coordinate monitoring, prediction,
and management activities (Jian and Jayas, 2024; Hurmat et al., 2026).
Although digital twins are novel, they may improve adaptive manage-
ment and the performance of storage systems (Ding and Gao, 2026).
5.7 Digital traceability
Blockchain technology and other forms of distributed ledgers have
the potential to record and share information on grain origin, storage
conditions, quality evaluation, treatment, and distribution throughout
the supply chain (Bosona and Gebresenbet, 2023). Increased traceability
FIGURE 4
Digital transformation pathways for stored grain management.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 11 frontiersin.org
promotes transparency and accountability, and these factors play a sig-
nificant role in ensuring the reliability of information on food safety
verification and quality assurance. Furthermore, the integration with IoT
sensors can enable the continuous recording of storage conditions and
quality information with the creation of traceable digital records
throughout the distribution and storage period (Rashed et al., 2025).
This can allow verification of handling practices, treatment history, and
grain quality at various stages of the supply chain. The use of blockchain
can also contribute to the exchange of information between producers,
storage entities, processors, and regulators; however, adoption might be
hindered by system incompatibility, the quality of information, and
transaction costs (Agarwal et al., 2024).
5.8 Technology readiness and
representative applications
Digital technologies in stored grain management are evolving
from laboratory studies toward pilot- and commercial-scale
applications, while the maturity of specific technologies varies consid-
erably (Schmidt et al., 2024). Constraints on broader adoption include
field validation, data generalizability, sensor accuracy, compatibility,
and cost (Y ang et al., 2026; Table 3).
6 Sustainable and reduced-risk
protection strategies
Sustainable stored grain protection aims to reduce reliance on
conventional synthetic insecticides and fumigants while maintaining
effective pest control and protecting grain quality, human health, and
the environment ( Rajashekar et al., 2025 ; Y ousuf et al., 2026b ).
Biological, physical, atmospheric, and behavioral approaches offer
complementary modes of action and can be selected according to
pest pressure, grain condition, storage environment, cost, scalability,
and management capacity (Guru et al., 2022).
TABLE 3 Digital technologies for stored grain management.
Technology Primary data
source
Function Advantages Limitations Development
stage
Literature
cited
Temperature/RH
sensors
Storage
environment
Environmental
monitoring; risk
assessment
Low cost; continuous
data
Drift; calibration Commercial
Coradi et al. (2022);
Crépon et al. (2023)
Moisture sensors Grain moisture
Moisture-risk
detection
Direct condition
indicator
Spatial variability;
calibration
Commercial/emerging Coradi et al. (2022)
CO2 sensing Biological activity
Early
deterioration/
insect detection
Non-destructive;
continuous
Non-specific signal;
background
respiration
Emerging/pilot
Ramachandran
(2022)
VOC sensors /
electronic nose
Grain/insect/
fungal VOCs
Early infestation/
fungal detection
Rapid; non-
destructive
VOCs complexity;
interference; drift;
calibration
Emerging/pilot
Liu et al. (2024);
Hou et al. (2024);
Borowik et al.
(2024); Mei et al.
(2025)
Acoustic sensing Insect activity
Early infestation
detection
Detects hidden
activity
Noise; calibration Pilot
Balingbing et al.
(2024b); Balingbing
et al. (2024a)
RGB imaging Grain/pest images
Classification;
grading
Rapid; accessible
Lighting; image
variability
Emerging/commercial
Santhanambika and
Maheswari (2025)
Hyperspectral/NIR
imaging
Spectral grain
response
Quality/
contamination
assessment
Non-destructive; rich
information
Cost; processing
complexity
Pilot/commercial
Johnson (2020);
Crépon et al. (2023)
IoT networks
Multiple sensor
streams
Real-time
monitoring; alerts
Remote access;
connectivity
Power; connectivity Commercial/emerging Crépon et al. (2023)
ML models
Sensor/image/
acoustic data
Detection;
classification
Handles nonlinear
patterns
Dataset dependence;
generalization
Emerging/pilot
Balingbing et al.
(2024b); Tian et al.
(2025)
Predictive models
Environmental +
historical data
Risk forecasting
Preventive
management
Validation;
transferability
Pilot/emerging Coradi et al. (2022)
Digital twins
Storage-system
data
Simulation;
optimization
Scenario testing
Data/computational
requirements
Emerging Dyck et al. (2023)
Blockchain systems
Traceability/
supply-chain data
Traceability;
transparency
Data integrity
Limited direct
storage benefit
Emerging
Feng et al. (2020);
Rejeb et al. (2026)
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 12 frontiersin.org
6.1 Botanical insecticides
Plant-derived materials such as essential oils, extracts, and pow-
ders have proven to be effective as insecticides, repellents, antifeed-
ants, and growth regulators against stored grain pests (Gautam and
Kumar, 2026; Usman et al., 2026). They are biodegradable and possess
various modes of action, making them attractive components of
reduced-risk management. However, the effectiveness of botanicals is
often limited by their chemical inconsistency, volatility, instability,
inconsistent application, and unfavorable odor, as well as high produc-
tion costs (Saikumar et al., 2025).
6.2 Microbial biopesticides
Microbial control agents, including Bacillus thuringiensis (Bt),
entomopathogenic fungi, and pest-specific viruses, offer alternative
means of pest control with specificity and reduced chemical input
compared to conventional pest management practices (Al-Jaddawi et
al., 2024; Murugesan et al., 2024). The effectiveness of these microbial
agents depends on several factors, including host range, environmen-
tal conditions, formulation, mode of application, and stability during
storage. Enhancing formulation stability and confirming their efficacy
under various environmental conditions still pose a challenge for their
large-scale use (Teixidó et al., 2022).
6.3 DEs and inert dusts
DEs and other inert dusts provide physical control by abrading the
cuticle of insect pests, resulting in loss of body fluid and mortality
(Chandaragi and Loganathan, 2025 ). They have proven efficacy
against several stored grain insect pests and offer advantages over con-
ventional chemical toxicants, and therefore are valuable components
of reduced-risk IPM ( Mortazavi et al., 2025 ; Wakil et al., 2025 ).
However, their efficacy is significantly reduced under conditions of
high humidity, and their application may negatively impact grain bulk
density, flowability, ease of handling, and processing ( Wakil et al.,
2023; Bashloo et al., 2026).
6.4 Hermetic storage and moisture
management
Hermetic storage involves the use of sufficiently sealed bags,
cocoons, and silos to restrict or eliminate the flow of air into and out
of the storage, thereby creating a low-O2 and high-CO2 environment
that suppresses insect activity (Navarro et al., 2024; Okparavero et al.,
2024). Its efficacy relies on the seal integrity, grain moisture, and the
initial quality of the grain before storage. Although hermetic storage
is considered an effective method against pest infestation, it does not
remove excessive moisture from the grain; thus, storing grains at
inappropriate moisture levels can still lead to spoilage ( Odjo et
al., 2025).
Elevated grain MC can provide increased water activity for micro-
bial growth. This can create an ideal environment for fungi and other
aerobic microbes to grow before a low-oxygen environment is estab-
lished (Ngoma et al., 2024). Although oxygen depletion in hermetic stor-
age inhibits fungal growth, high-moisture grains can lead to anaerobic
fermentation, which can cause undesired changes in grain quality and
decrease seed viability (Y ewle et al., 2024). Additionally, temperature
differences within bulk grain storage can cause air movement and mois-
ture migration, resulting in moisture buildup within specific areas and,
therefore, condensation (Yin et al., 2025a). Safe moisture levels should
therefore be determined for each grain species, storage temperature, and
storage duration (Kamel et al., 2024).
For moisture regulation, desiccant-based methods such as silica gel,
activated alumina, zeolite, and calcium chloride can be used, provided
that they are appropriately selected and implemented (Y ousuf et al.,
2026a; Darfour et al., 2026). However, their use requires optimization of
the dose, capacity to absorb moisture, location, and storage time to avoid
excessive dehydration, adverse effects on grain quality, contamination,
and high cost (Sultana et al., 2021).
Combining efficient drying with hermetic storage can therefore offer
complementary protection against insect infestation and moisture-
related deterioration (Adase et al., 2026). Recent evidence from field
experiments on wheat storage in Pakistan showed differences in pest
infestation and grain quality performance between conventional and
hermetic storage facilities under ambient conditions, highlighting the
crucial role of the storage structure for long-term grain protection
(Y ousuf et al., 2025). Likewise, evidence from smallholder systems also
highlights the practical value of hermetic storage. In a randomized
encouragement study among 692 maize-growing households in Nepal,
the provision of PICS hermetic bags increased maize storage and, among
farmers offered a full subsidy, reduced post-harvest storage losses by
10 kg, corresponding to a 32.92% reduction (Nepali and Maharjan, 2025).
6.5 Modified and controlled atmospheres
Modified and controlled atmospheres systems are deliberately modi-
fied to alter O2 and CO2 contents (Navarro et al., 2024; Lin et al., 2025).
In certain instances, inert gases such as nitrogen can be introduced to
suffocate pests and prevent microbial proliferation ( Navarro and
Navarro, 2020). Unlike passive hermetic storage, such systems generally
require controlled gas delivery, gas monitoring, and specific infrastruc-
ture. Thus, their viability is determined by the storage capacity, equip-
ment, operating costs, and technical capabilities (Guru et al., 2022).
6.6 Physical control
Physical control methods can eliminate storage pests without leaving
traditional pesticide residues (Kopacki et al., 2021). Heat treatments can
be used to disinfect either grain or the storage itself by exposing insects
to a lethal temperature, while artificial cooling technologies allow for the
inhibition of insect development and other biochemical processes lead-
ing to grain deterioration (Mason and Strait, 2019; Navarro and Navarro,
2020). The main constraints of this method are energy costs, equipment
availability, uniformity of treatment, exposure time, and quality of grain
after treatment (Ziegler et al., 2021).
6.7 Ozone and other emerging approaches
Ozone (O3) appears to be a promising tool for stored grain protec-
tion because of its oxidizing properties against insect pests and microor-
ganisms. However, its effectiveness depends on ozone concentration,
exposure time, and distribution within the grain mass (Sitoe et al., 2025).
Inadequate penetration or uneven gas distribution may reduce treatment
efficacy, particularly in large-scale storage systems. Moreover, researchers
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 13 frontiersin.org
should also consider alternative approaches, including cold plasma and
ultraviolet-based techniques, to control insect infestation and microbial
growth in stored products (Eshtiaghi et al., 2025). Although these tech-
nologies show potential, their efficacy in terms of treatment, grain qual-
ity, worker safety, energy requirements, treatment consistency, scalability,
and economic feasibility must be evaluated further before adoption (Lin
et al., 2025).
6.8 Semiochemicals and behavioral control
Semiochemicals, including pheromones and additional behavior-
modifying compounds, can be used for pest monitoring and manage-
ment using mass trapping, attract-and-kill, or push-pull approaches
(Gong et al., 2026). Their effects on non-target species are limited by
their high specificity, and these approaches can complement other pest
management practices (Adamič Zamljen et al., 2025). Their efficacy
depends on pest biology, trap design, deployment strategy, and target
pest population densities. Therefore, they need to be designed and imple-
mented as part of integrated programs that include monitoring and other
IPM practices (Mweresa et al., 2020; Athanassiou et al., 2025a).
6.9 Integrated reduced-risk pest
management
The performance of individual reduced-risk technologies can
be enhanced through their integration with monitoring and deci -
sion-support systems ( Schmidt et al., 2023 ; Lisboa et al., 2024 ).
Digital IPM incorporates environmental and pest monitoring data
with predictive algorithms, predefined risk thresholds, and tar -
geted intervention ( Bal et al., 2026 ). Sensor-based detection com -
bined with AI-assisted predictions can help detect emerging risks
and suggest when and what type of intervention is needed ( Kępka-
Borkowska et al., 2025 ; Matsiko et al., 2026 ), while hermetic
storage, biological control, inert dusts, botanicals, or other
reduced risk measures may be applied depending on the detected
risk and specific storage conditions ( Y ousuf et al., 2026b ). This
approach shifts the emphasis of IPM from routine treatments to
early intervention based on specific risk and more efficient use of
available control options ( Guru et al., 2022 ). Figure 5 conceptually
illustrates the hierarchy of sustainable and reduced-risk protection
strategies discussed in this section and their potential inclusion in
digital IPM and sustainability assessment.
6.10 Environmental and economic
implications
Sustainable protection strategies should be evaluated not only
by pest-control efficacy but also by environmental impact, cost, scal-
ability, grain compatibility, labor requirements, energy demand, and
reliability under different climatic conditions ( Rajashekar et al.,
2025). While reduced reliance on conventional chemicals has the
potential to lower residues and ecological concerns, alternative tech-
nologies can pose increased material, energy, labor, or monitoring
costs and variable efficacy across storage conditions ( Lisboa et al.,
2024). Comparative assessment is required to identify effective com-
binations that are technically and economically viable for specific
commodities and storage systems.
Previous studies have reported various positive impacts of using
advanced grain management technologies, including a decrease in
grain weight loss, grain damage, and insect infestation in hermetic
storage, as well as improvement in terms of pest monitoring and con-
trol and other climate-related variables through the use of digital and
climate-controlled technologies (Tian et al., 2025; Sun et al., 2026) as
illustrated in Figure 6. Panel A shows the reduction in grain damage
and weight loss compared to the control or conventional storage con-
ditions. Panel B reports the detection and classification accuracy of
each study using AI and sensors compared to reference observations.
FIGURE 5
Hierarchy of sustainable and reduced-risk pest management.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 14 frontiersin.org
Thus, the percentages in Panel B show actual results rather than rela-
tive improvements over traditional approaches. Therefore, results
cannot be compared directly between the two panels.
6.11 Life-cycle assessment
Life-cycle assessment (LCA) offers a systematic approach to
compare the environmental impacts of conventional and reduced-
risk storage strategies by taking into account the impacts of rel -
evant life cycles of the technology involved ( Ma and Rosentrater,
2019). Depending on the system boundary, the assessment may
encompass production, transport, operation, maintenance, and
end-of-life, with indicators that may include energy use, green -
house gas emissions, material consumption, waste generated, and
other environmental impact categories ( Lisboa et al., 2024 ). For
stored grain protection, LCA can demonstrate trade-offs that may
not be evident from the efficacy of pest control or the reduction
in storage losses ( Payandeh et al., 2021 ). For example, Ignacio et
al. (2023) conducted a cradle-to-grave LCA of six commercial
hermetic storage bags for maize and reported that bag manufac -
turing constituted approximately 84.6–90.8% of the total environ -
mental impact, whereas the carbon footprints ranged from 1.1 to
1.7 kg CO 2-equivalent per bag during the considered period of
service life. Their findings indicate that the reduced losses associ -
ated with hermetic technologies do not necessarily eliminate their
environmental burdens related to the storage technology itself.
Likewise, de Sousa et al. (2023) emphasized the importance of
including life cycle and cost analyses when assessing environmen -
tally-friendly stored-rice pest management solutions. Hence, a
combination of LCA with technical and economic assessment can
help to differentiate between technologies that only reduce chemi -
cal inputs and those that have an overall environmental benefit
during their life cycle and contribute to comprehensive sustain -
ability decisions.
7 Integrated and climate-resilient grain
storage
The multiple risks associated with stored grains are interrelated,
and individual measures offer only limited protection. Efficient grain
management thus demands a coordinated strategy that integrates bio-
logical knowledge, environmental monitoring, predictive analytics,
climate information, and intervention factors within a single decision-
support system (Schmidt et al., 2023).
7.1 Predictive framework for grain
management
Unlike conventional grain management frameworks, predictive
grain management can be conceptualized as a closed-loop system
encompassing sensing, analysis, prediction, decision-making, inter-
vention, and verification (Jayavigneshwaran and Dishri, 2026; Latha,
2026). These six stages are selected according to their functional roles
in connecting information collection, risk assessment, management
decisions, and intervention outcomes. These constructs are organized
as an interconnected process in which verification provides feedback
to improve subsequent monitoring and management decisions. In this
system, variables related to grain, storage environment, and pests are
constantly monitored using sensors and detection techniques
(Badgujar et al., 2026). The collected information is evaluated in con-
junction with historical and contextual data to detect potential issues
and trends (Y ang et al., 2026). Thereafter, predictive analytics estimate
the likelihood and timing of pest invasion, fungal proliferation, mois-
ture-related deterioration, and other hazards in storage. Predictions
from these models are then translated into decisions based on biologi-
cal, environmental, economic, and sustainability thresholds. Following
that, integrated interventions are implemented in accordance with
site-specific and context-specific priorities, with an emphasis on
reduced-risk solutions (Jeur et al., 2025 ; Luo et al., 2025 ). Finally,
FIGURE 6
Literature-derived physical grain protection and AI-enabled pest monitoring outcomes. Reported values represent study-specific grain weight-loss and
damage reductions, temperature regulation, and AI- and sensor-based detection performance. These outcomes involve different metrics and
experimental conditions and are not directly comparable across technologies (Tola et al., 2020; Coradi et al., 2022; Rigopoulou et al., 2023; Xu et al.,
2023; Balingbing et al., 2024b; Liu et al., 2024; Ngoma et al., 2024; Tian et al., 2025; Sun et al., 2026; Yousuf et al., 2026a,b; Badgujar et al., 2026; Gong
et al., 2026).
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 15 frontiersin.org
storage conditions and intervention measures are re-examined to
facilitate the refinement of management decisions and predictive
models (Garima et al., 2024 ). Such a closed-loop system is distin -
guished from conventional systems in which monitoring and inter-
vention are often disconnected (Figure 7).
7.2 Digital IPM
Digital IPM applies the above-described framework to the domain
of pest management involving IoT-based approaches for analyzing
and predicting pest problems, taking into account the factors and
measures appropriate to the local environment (Ghosh and Markad,
2023; Aiwale et al., 2025). It aims to reduce dependence on regular
inspections and treatments by continuously updating risk assessments
and recommending timely interventions for effective control
(Badgujar et al., 2026). The success of such IPM depends on the qual-
ity of information provided by sensors, the accuracy of predictive
models, the selection of appropriate thresholds, and the availability of
effective measures.
7.3 Climate-responsive decision-making
Climate-responsive grain management practices consider
both current storage conditions and potential changes during the
decision-making process ( Poetelea et al., 2026 ). Real-time obser -
vations can be combined with weather forecasts, seasonal climate
data, and biological response models to identify periods of
enhanced risk ( Rodrigues et al., 2024 ). This information could be
used to adjust the level of surveillance, aeration, cooling, moisture
control, and pest-control measures ( Scheff and Phillips, 2022 ; Qu
et al., 2023 ). The objective is to anticipate climate-related storage
risks and implement preventive measures before significant grain
deterioration occurs.
7.4 Risk-based intervention
One of the significant advantages of predictive management
includes the possibility of replacing routine or blanket treatment with
risk-based interventions (Wang et al., 2025). The intensity of manage-
ment can be aligned with the likelihood and possible consequences of
deterioration. For instance, low-risk situations may require continued
surveillance, whereas higher risk may require intensified measures,
including targeted surveillance, prevention, or intensive control, based
on defined thresholds and management objectives ( Li et al., 2020 ;
Chen et al., 2026). A deep-learning framework was recently proposed
by Li X. et al. (2025), who classified grain storage conditions and pre-
dicted changes in the 3D grain-temperature field, thus enabling early
warning of possible risks and corresponding measures. These
approaches may help reduce unnecessary treatments, optimize
resource utilization, and promote more effective use of grain protec-
tion measures.
7.5 Commodity-specific management
Predictive management should consider the variation among
stored grains in relation to their physical properties, storage behavior,
moisture relations, pest complexes, end uses, and market require -
ments (Dasore et al., 2025; Latha, 2026). These properties may have
an influence on the adaptation and deterioration processes, which
suggests that predictive models developed for one category may not
necessarily be applicable or effective for other products without appro-
priate adaptation and validation ( Anukiruthika and Jayas, 2024 ;
FIGURE 7
Integrated framework for predictive and climate-resilient grain management.
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 16 frontiersin.org
Zhang L. et al., 2026 ). Thus, it is necessary to develop appropriate
databases, threshold sets, predictive models, and interventions for dif-
ferent commodities to ensure the accuracy, transferability, and prac-
ticality of the models in diversified storage systems.
7.6 Smallholder and industrial storage
systems
The implementation of smart grain management depends on vari-
ous key factors, such as the scale of storage and the infrastructure of
the storage facility. Industrial enterprises can manage their commodi-
ties using interconnected sensors, automatic aeration and cooling,
cloud platforms, analytics software, and digital twin technologies
(Rodrigues et al., 2024 ). Conversely, small-scale producers mostly
adopt simple and resilient approaches, i.e., hermetic storage, afford-
able sensors, smartphone applications, warning systems, and local
data storage (Y ousuf et al., 2025). Therefore, developing technologies
that can be used at industrial, community, and individual levels is
critical for digital and climate-smart grain storage ( Alharbawee et
al., 2026).
7.7 Economic feasibility and scalability
The technical performance of a given approach is not the sole
determinant of its adoption. The cost of installation, operation
and maintenance, energy inputs, connectivity, technical expertise,
loss reduction, and return on investment are some of the key con -
siderations that play a major role in adopting smart and sustain -
able grain storage ( Schmidt et al., 2024 ; Oyewole et al., 2025 ).
Although expensive technologies are applicable in large-scale
commercial grain storage, such solutions are inaccessible in the
smallholder farming system. Thus, developing scalable agricul -
tural practices that consider cost–benefit dynamics, technical
compatibility, ease of maintenance, and the availability of spares
and accessories is key to achieving wider adoption ( Ziegler et al.,
2021; Kenkel, 2022 ). Public-private partnerships, cooperatives,
financing programs, and targeted policies could facilitate the
required investment in adoption-supporting infrastructure
(Gosa, 2026 ).
Evidence from Ethiopia provides an illustration of why technical
performance and socioeconomic outcomes need to be considered
separately. In a randomized controlled trial among 871 smallholder
maize farmers, access to PICS hermetic bags resulted in only a mar-
ginal extension of storage duration and had no detectable impact on
household welfare despite significant seasonal price variation. The
authors associated the limited welfare effect partly with relatively low
storage losses in the study area and the prevalence of other pest-con-
trol practices (Negede et al., 2024).
7.8 Pathway to food-security outcomes
The ultimate value of integrated grain management is determined
by its ability to improve food-system outcomes. Improved early detec-
tion and prediction capacities can enable timely intervention, while
sustainable and climate-resilient technologies can reduce the magni-
tude and frequency of storage losses (Y ang et al., 2026). These inter-
ventions can enhance grain availability and quality, improve food
safety and marketability, build resilience in storage systems, and
potentially increase farmer revenues (Baidhe et al., 2024). In this con-
text, the management of stored grain is considered an essential com-
ponent of resilient and sustainable food systems ( AlFadhly et
al., 2024).
8 Limitations and research gaps
The transition to sustainable, smart, and predictive stored-grain
management is currently hindered by various technical, biological,
economic, and institutional factors (Schmidt et al., 2024). Although
emerging digital sensing, AI, and reduced-risk technologies have the
potential to revolutionize grain storage management, their effective
use requires accessible data, advanced sensing and modeling, biologi-
cal knowledge, cost-effectiveness, and supportive policies
(Anukiruthika and Jayas, 2024). In addition, their limited validation
in the field and the absence of standardized frameworks undermine
their ability to be applied at different scales and storage systems.
Smallholders face additional challenges associated with affordability,
digital literacy, and infrastructure, while regulatory uncertainty hin-
ders the transition of promising technologies from research to practice
(Adil et al., 2026).
8.1 Data limitations
The reliability and diversity of datasets play an important role in
predictive stored-grain management. Many of the existing datasets are
restricted in their application due to the influence of regional distribu-
tion, commodities, controlled environments, and inconsistent proto-
cols, limiting the use of the models in diverse storage systems and
climates (Wu et al., 2023). There is a lack of standardized, long-term
data repositories for multiple commodities, limiting the development,
training, and validation of models that can be used in varying envi-
ronmental and operational conditions (Li et al., 2026). Therefore, it is
necessary to create coordinated databases with consistent measure-
ments across different commodities, storage conditions, and climates
that will improve model development and independent validation.
Furthermore, data obtained from different sources like AI sensors,
IoT platforms, and laboratory experiments is often heterogeneous,
making data integration and the development of unified large-scale
models challenging. In most cases, these models are tested only in
pilot or laboratory settings, which decreases their efficiency and accu-
racy under field conditions (Athanassiou et al., 2025a; Kumar et al.,
2026). Recent research emphasizes the necessity of creating open
repositories with standardized data collection and reporting protocols
to compare the results from different regions and increase the gener-
alizability of the AI models (Anukiruthika and Jayas, 2025; Schmidt
et al., 2025). Without these measures, scaling predictive solutions for
grain-storage systems and integrating them globally will remain
challenging.
8.2 Limitations of AI and sensor
technologies
The quality and representativeness of training datasets are critical
concerns for the effectiveness of AI and ML methods (Barbedo, 2025).
The models that have been developed for use in a specific grain or
environment may not be applicable in different regions or under
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 17 frontiersin.org
different storage conditions, and their limited interpretability and
uncertainty can erode user confidence ( Yunshandan et al., 2025 ).
Recent studies exploring the use of AI in grain storage that have
focused on detecting pests, determining quality, and predicting out-
comes demonstrate the significance of these technologies in address-
ing issues of data quality, prediction accuracy, and regulatory
compliance in order to facilitate their practical adoption (Anukiruthika
and Jayas, 2024; Zhang Y . et al., 2025). Hence, there is a need for fur-
ther research to support cross-environment validation, quantify
model uncertainty, and develop interpretable AI models for sustain-
able grain storage (Pai et al., 2025).
Beyond the limitations of AI models, the reliability of predictive
grain management also depends on the accuracy and operational
stability of the sensors used for data collection. To ensure high accu-
racy in continuous monitoring systems, it is critical to use sensors
that remain accurate and operational in different storage conditions.
Many aspects, including calibration, exposure to environmental
influences, sensor durability, power requirements, and spatial varia-
tion of stored grain, could impact the accuracy of data acquisition
(Coradi et al., 2022). Sensor positioning and spacing are also impor-
tant factors because local changes in temperature or moisture might
go undetected with insufficient monitoring points (Wu et al., 2023).
The accuracy of CO2 and VOCs measurements in bulk grain storage
is affected by sensor location, gas diffusion, and airflow through the
inter-grain space. Sensors positioned farther from the affected areas
may fail to detect localized spoilage hotspots ( Nunes et al., 2023 ).
Therefore, using a well-positioned sensor arrangement or multi-point
gas sampling is important for improving hotspot detection. Dust
build-up can cause blocking of inlet ports and affect sensor response,
requiring a suitable enclosure with dust filtering and cleaning capa-
bility that will not restrict gas flow (Crépon et al., 2023). Sensor drift
can occur due to sensor aging, humidity fluctuations, and exposure
to volatile compounds, necessitating periodic calibration using
appropriate reference standards, baseline checks, and drift correction
(Badgujar et al., 2026).
The long-term operation of these monitoring systems could be
limited by the need to recharge batteries, communication breakdowns,
and the influence of environmental variables, necessitating regular
calibration and maintenance (Jain et al., 2026). In addition, research
on continuous grain monitoring using IoT technology contributes
insights into the advantages of continuous monitoring while empha-
sizing the importance of addressing the issues of sensor configura-
tions, data acquisition principles, and the validation of results in
diverse storage conditions (Crépon et al., 2023). Therefore, the imple-
mentation of such technology on a larger scale requires its rational
design and placement, as well as improvements in sensor characteris-
tics, including their power supply, calibration, and maintenance.
8.3 Interoperability and standardization
Integration of various components involved in grain storage,
including sensors, IoT platforms, databases, and AI systems, remains
constrained by disparities in hardware, communications, data formats,
and proprietary platforms. Such limitations may restrict data exchange
and impair the progress of integrated decision-support systems
(Anukiruthika and Jayas, 2025). According to the recent reviews on
smart grain-storage technologies, the key emerging requirements
include considerations of interoperability and integration for connect-
ing IoT, AI, digital twins, and other advanced digital technologies
(Anukiruthika and Jayas, 2024). As a result, there is a need for stan-
dardization in data acquisition, communications, and model exchange
to enable interoperability and scalability of grain management
systems.
8.4 Biological complexity
Stored grain ecosystems consist of complex biological interac -
tions among pests, fungi, grain properties, moisture, temperature,
and other related variables ( Jian and Jayas, 2012). Such associations
can be exceptionally complicated and challenging because biologi -
cal processes can affect both grain condition and the storage envi -
ronment, while fungal development as well as insect activity may
have interacting effects over time ( Hernandez Nopsa et al., 2015 ).
Previous models have made substantial contributions by providing
a better understanding of pest biology, fungal growth, and the
properties of the stored grain, as well as the importance of consid -
ering various interconnected physical, biological, and economic
processes in mechanistic models ( Y ousuf et al., 2026b). In the con-
text of recent contributions to multifactorial modeling of fungal
growth, it becomes apparent that temperature and moisture con -
siderations in biological prediction and control models are of par -
ticular importance ( Li D. et al., 2025 ). Therefore, more
comprehensive integration of biological processes is required to
achieve a more profound understanding of the risks involved in real
storage conditions.
8.5 Economic and infrastructure barriers
The implementation of smart storage technologies is influenced
by the optimization of investment, maintenance, and operating costs,
loss reduction, and profitability (Chowdhury et al., 2021). Although
advanced sensing, connectivity, and AI infrastructure are applicable
in commercial storage facilities, these technologies are financially
impractical in smaller systems. Moreover, the availability of electric-
ity, maintenance, and technical infrastructure may limit the spread
of innovative storage technologies (Amiri, 2026). Recent studies indi-
cate that the adoption and economic performance of advanced stor-
age solutions are influenced by costs, training, financing, storage
scale, and anticipated economic results ( Mbesa et al., 2025 ;
Ngwenyama et al., 2026 ). Therefore, economic evaluation should
accompany the technical assessment to ensure that the proposed
technologies have economic viability and technical feasibility.
8.6 Smallholder adoption
The adoption rate of improved storage technologies by small -
holders is determined by cost-effectiveness, technical knowledge,
financing, local availability, and the adaptability of the technology
to the current storage practices ( Manandhar et al., 2018 ). Recent
evidence from Malawi identified perceived high cost, limited avail -
ability, insufficient knowledge, and low yields as important barriers
to continued use of hermetic bags ( Nakoma-Ngoma et al., 2025b ).
A broader assessment across Sub-Saharan Africa and South Asia
similarly identifies training, local availability, financing, and
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 18 frontiersin.org
institutional support as important determinants of post-harvest
technology adoption ( Nakoma-Ngoma et al., 2025a ). The develop -
ment of simple, low-power, offline-capable, and locally appropriate
solutions is therefore important for broader and more equitable
adoption.
In addition to the above-mentioned factors, gender-related dif-
ferences, cultural preferences, and trust in extension services also
play important roles in the adoption of improved technologies.
Studies have shown that women farmers, as the primary caretakers
of the household grain, have less access to training and finance
(Ndosi et al., 2026). At the same time, the lack of proper distribution
channels and after-sales services undermines farmers’ confidence,
leading to low adoption rates (Schmidt et al., 2024). Therefore, pro-
moting farmer cooperatives, providing training at institutional levels,
and ensuring access to finance are viable options for promoting
adoption.
8.7 Policy and regulatory challenges
Wider adoption is also influenced by regulatory requirements
governing AI, data management, pesticide use, food safety, and tech-
nology certification ( Gong M. et al., 2025 ). Recent assessments of
AI-based grain management identify data privacy and regulatory
concerns as key challenges in implementation, which differ consider-
ably across countries or regions according to the level of legal matu-
rity, enforcement, and institutional interests (Anukiruthika and Jayas,
2025). At the agricultural level, weak institutional support and frag-
mented policy environments can also hinder the dissemination of
post-harvest technologies (Nakoma-Ngoma et al., 2025b).
In addition, the regulatory requirements may also differ for
each jurisdiction in terms of pesticide authorizations, residue
limits, and food-safety standards. For example, the European
Union (EU) implements the General Data Protection Regulation
for personal data processing and sets strict requirements for pesti -
cide approval and maximum residue levels ( FAO, 2024 ; EU
(European Commission), 2026 ). At the international level, Codex
standards offer the basis for harmonizing food-safety requirements,
but their effective implementation and enforcement require effec -
tive national regulatory and institutional frameworks ( FAO, 2025).
Diverse certification requirements and the variation in food-safety
control practices reduce the diffusion of innovations across borders
(Anukiruthika and Jayas, 2025 ; Gong M. et al., 2025 ). Regulatory
frameworks should therefore support innovation while maintain -
ing food-safety, environmental, and responsible-technology
standards.
8.8 Limited incentives for sustainable
technologies
Reduced-risk technologies may face weak short-term economic
incentives despite potential long-term benefits. Adoption of
advanced technologies can be limited by a tendency toward the
more familiar and immediately accessible options (Bokusheva et al.,
2012). Recent evidence demonstrates that policy design and finan -
cial incentives can have a significant impact on adoption: a partial
subsidy increased hermetic bag adoption sixfold in an Ethiopian
field experiment, and delayed payment increased willingness to pay
for hermetic storage among smallholders in another study (Brander
et al., 2025 ; Mbesa et al., 2025 ). Evidence from Tanzania also
suggests that training, income, market orientation, and economic
returns can all influence hermetic storage adoption ( Mbesa et al.,
2026). This evidence suggests that targeted investments, market
incentives, training, and policy interventions may be effective in
promoting the adoption of cost-effective sustainable technologies.
8.9 Research gap matrix
The major impediments and research priorities mentioned above
have been grouped in Table 4. The matrix differentiates between
knowledge gaps and implementation barriers and highlights the areas
of emphasis for more reliable, scalable, and inclusive stored grain
management.
Overall, the primary challenge is not the absence of individual
technologies but their adequate integration across various storage sys-
tems. Addressing this challenge requires improvements in data quality,
biological modeling, sensing, interoperability, economic feasibility,
and supportive policies.
9 Future directions
Emerging stored grain management practices are likely to
progress from isolated monitoring and control technologies
toward more connected, predictive, adaptive, and automated
approaches ( Madhu, 2025 ). Future research efforts should be
directed toward the development of technologies that can operate
effectively under different storage and climatic conditions while
incorporating sensing, analytics, prediction, and control mecha -
nisms that support increasingly adaptive management systems
(Nath et al., 2024 ).
9.1 Edge intelligence and low-cost sensor
networks
Next-generation monitoring technologies should integrate low-
cost wireless sensors with edge AI to allow for local data processing
(Gong R. et al., 2025). The utilization of durable and energy-efficient
technologies that can measure temperature, humidity, grain moisture,
gases, acoustics, and spectral properties is expected to ensure the
monitoring of multiple storage units (Suhra et al., 2025). Edge AI has
the potential to limit the need for a permanent connection to the
cloud while enabling faster data processing when there is no internet
connectivity (Olatinwo et al., 2025; Samanta and Saha, 2026). Future
technologies should focus on reducing the power consumption of
monitoring instruments, as well as improving their robustness, com-
patibility, and applicability.
9.2 Multimodal sensing and predictive AI
Multimodal sensing enables the combination of different signals,
including temperature, moisture, CO2/O2, acoustic activity, imaging,
and spectral signature data, into predictive models ( Zhou et al.,
2026). The resulting models can provide a more comprehensive
understanding of the monitored environment and biological activity,
with greater robustness across grain types, climates, and storage sys-
tems (Schmidt et al., 2023; Lisboa et al., 2024). Combining such pre-
dictive systems with interpretable AI could enable the clarification of
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 19 frontiersin.org
mechanistic relationships while providing an assessment of uncer-
tainties that could be involved in the prediction of specific risks
(Abou Samra and Ali, 2026). In addition, digital twin technologies
can enable alternative scenario testing, while the incorporation of
predicted weather and seasonal data can ensure adaptation to evolv-
ing conditions (Padhiary et al., 2026; Wu and Fu, 2026).
9.3 Precision pest management
Predictive analytics facilitates targeted pest control, where mea-
sures are taken based on anticipated risks, infestation severity, and
economic importance (Bal et al., 2026). Contemporary approaches,
including smart traps, biosensors, and early-warning technologies,
can allow timing and placement of interventions to optimize control
and minimize excessive pesticide applications, thus enhancing the
efficacy of integrated sustainable management practices (Divya et al.,
2026). In addition, innovative nanotechnologies such as nano-coat-
ings, formulations, and biosensors have the potential to transform
prevailing practices to improve detection and protection against
pests; however, their performance, safety, ecological consequences,
regulations, and costs need to be scrutinized prior to wide adoption
(Singh et al., 2025).
9.4 Human-AI cooperative systems
Future developments and innovations in sensing, predictive models,
and automated control could allow storage systems to identify abnormal
conditions, predict consequences, suggest appropriate responses, and
modulate aeration or cooling (Jian and Jayas, 2024; Roy, 2026). Although
high levels of autonomy can be viable in automated facilities, human-AI
collaboration is likely to persist in diverse storage contexts (Agrawal et
al., 2025; Ayoub, 2026). In such systems, AI can be utilized to offer con-
stant analysis and recommendations while also preserving managers’
contextual judgment and final decision-making power under conditions
of uncertainty (Dili and Ansong, 2025; Liu J. et al., 2025). Future research
should therefore focus on reliable human-AI interaction, safe automa-
tion, dealing with uncertainty, and mechanisms by which systems incor-
porate operational feedback.
Collectively, these advances suggest the emergence of stored
grain systems capable of learning from real-time and historical infor-
mation as well as adapting management responses to changing bio-
logical and environmental conditions. Advances toward this vision
will be based on the convergence of accurate sensing, transferable
predictive algorithms, sustainable pest management strategies, and
appropriate levels of automation in compatible systems. These devel-
opments have the potential to enable early risk detection and more
targeted and adaptive stored grain management in a variety of
contexts.
10 Conclusion
Grain storage systems face various interrelated challenges, including
pest infestation, fungal contamination, moisture and temperature insta-
bility, climatic variations, and food safety risks. The challenges are exac-
erbated by conventional management approaches that often rely on
periodic inspections and reactive intervention after deterioration
TABLE 4 Research gaps in smart stored grain management.
Research gap Current status Major consequence Literature cited
Standardized multi-grain
datasets
Fragmented and
geographically limited
Poor model transferability and
validation
Schmidt et al. (2023); Anukiruthika and Jayas (2024); Balingbing et
al. (2024a)
Field validation of AI
models
Mostly laboratory/pilot based Uncertain real-world reliability
Lutz and Coradi (2022); Tian et al. (2025); Zhu et al. (2025); Zhang
X. et al. (2026)
Pest-fungus predictive
modeling
Biological interactions
oversimplified
Incomplete risk prediction
Okori et al. (2022); Stathas et al. (2023); Wang et al. (2025); Y ousuf et
al. (2026b)
Smallholder affordability
Advanced systems remain
costly
Unequal technology adoption
Manandhar et al. (2018); Baributsa and Concepcion Ignacio (2020);
Nakoma-Ngoma et al. (2025a); Jain et al. (2026)
Sensor reliability and
standardization
Variable performance across
environments
Inconsistent monitoring data Lutz and Coradi (2022); Crépon et al. (2023); Zhang L. et al. (2026)
Multimodal sensing
Emerging but limited field
integration
Incomplete risk representation Fuentes-Peñailillo et al. (2024); Lisboa et al. (2024)
Interoperability standards
Fragmented platforms and
protocols
Difficult system integration Schmidt et al. (2025); Rejeb et al. (2026)
Economic/ROI evidence
Limited long-term field
assessments
Uncertain investment decisions
Stathers et al. (2020); Ngoma et al. (2024); Nepali and Maharjan
(2025); Odjo et al. (2025)
AI explainability and
uncertainty
Still developing Limited user trust Pai et al. (2025); Abou Samra and Ali (2026); Zhang X. et al. (2026)
Policy and regulatory
frameworks
Developing and regionally
fragmented
Delayed adoption Stathers et al. (2020); Madhu (2025); Nakoma-Ngoma et al. (2025b);
Incentives for sustainable
technologies
Limited in many systems Low uptake despite benefits
Manandhar et al. (2018); Baributsa and Concepcion Ignacio (2020);
Nepali and Maharjan (2025)
Yousuf et al. 10.3389/fsufs.2026.2020632
Frontiers in Sustainable Food Systems 20 frontiersin.org
becomes apparent. Under changing climate conditions, the maintenance
of the quality and safety of the grain requires more anticipation and
adaptability.
Sustainable grain management approaches, including her -
metically sealed structures, biopesticides, botanicals, DEs, con -
trolled atmosphere, and various other lower-risk alternatives, can
reduce reliance on traditional chemical control. However, their
effectiveness depends on several factors, including storage condi -
tions, pest pressures, timing, and management techniques. The
combined use of these measures, together with AI- and IoT-based
sensing and prediction tools, can enable early detection and
assessment of risks and the implementation of targeted interven -
tions. The conceptual framework proposed in this review inte -
grates these elements to support stored-grain management while
accounting for complex biological and environmental factors.
Future developments should be based on standardized datasets,
cost-effective and high-quality sensing, biologically informed predic-
tive models, field validation, and an understanding of economic
viability and long-term sustainability. Relevant policies, financing
schemes, food safety regulations, and technologies adapted to diverse
storage environments will favor widespread adoption. Ultimately, a
multidisciplinary approach that includes biological, technological,
climatic, and socioeconomic aspects will lead to improved stored
grain management. Together, these factors, combined with the sus-
tainable use of pest control measures, will contribute to reducing
post-harvest losses, ensuring grain quality and safety, improving stor-
age capacity, and supporting sustainable food systems.
Author contributions
HY: Conceptualization, Writing – original draft, Writing – review
& editing, Investigation, Methodology, Visualization, Project admin-
istration. MY: Writing – review & editing, Supervision, Validation.
KR: Writing – review & editing, Supervision, Validation. MUY:
Writing – review & editing.
Funding
The author(s) declared that financial support was not received for
this work and/or its publication.
Acknowledgments
The authors acknowledge the contributions of researchers and
scientists whose published work formed the basis of this review. The
authors also acknowledge BioRender for providing the platform used
to create the figures presented in this review.
Conflict of interest
The author(s) declared that this work was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.
The author KR declared that they were an editorial board member
of Frontiers, at the time of submission. This had no impact on the peer
review process and the final decision.
Generative AI
论文原文与出版信息
Frontiers in Sustainable Food Systems