Using these novel tools to aid processors.
Almost all of my students talk to me about using artificial intelligence (AI) tools for various tasks at the university or on their internships. Indeed, some confess to using it to write their term papers, although they are embarrassed to admit.
Most, however, tell me about using it as an aid for simple tasks, such as: checking grammar and sentence structure of an essay, finding reference sources when writing papers, showing how to do complex calculations, and — one that I really think is useful — to summarize scientific papers into digestible, short content.
AI seems to be the biggest discussion point in many aspects of our industries this year. Have you heard about how AI is going to revolutionize the grain processing industries too? I am intrigued as to how this new technology is going to make our lives easier and help fill the labor shortage in many of our industries. Perhaps I’m a bit skeptical.
So, I thought I would play around with one AI platform, ChatGPT, to see if it could live up to the hype. Here are some examples of my findings.
Prompt: “Image of a concrete grain elevator.”
Result: See first image at right.
Prompt: “Rework with clean concrete surfaces, bright sunlight, instead of a head house use a steel tower. Make it look like it was constructed in 2025. Use a bright blue sky with white cumulus clouds.”
Result: See second image at right.
Author’s evaluation: “Overall, these images seem to me to be somewhat interesting, but not extremely helpful. And I’m not sure what the story is regarding the bridge to nowhere! What about using ChatGPT for other grain processing tasks?”
Prompt: “Top 5 challenges facing the grain storage and processing industries and potential solutions.”
Result:
1. Grain quality degradation (moisture, pest damage, spoilage)
Challenges
- High moisture leads to mold growth and mycotoxins
- Insect infestations reduce quality and market value
- Temperature fluctuations cause condensation and spoilage
Potential Solutions
- Advanced monitoring systems (IoT sensors for moisture, temperature, and carbon dioxide)
- Aeration management with automated fans and predictive control
- Regular fumigation and integrated pest management
- Grain turning and conditioning to equalize temperature
- Hermetic or sealed storage to reduce oxygen and limit pests
2. Infrastructure aging and maintenance costs
Challenges
- Many concrete elevators are several decades old
- Structural deterioration increases safety risks
- Repair and modernization are expensive and disruptive
Potential Solutions
- Routine structural health assessments using drones and 3D scanning
- Predictive maintenance schedules driven by sensors and inspections
- Upgrading to modular steel storage units where expansion is needed
- Seeking grants/financing for modernization (especially in rural regions)
- Implementing risk-based prioritization for repairs and upgrades
3. Supply chain disruptions and logistics bottlenecks
Challenges
- Transportation delays (rail, trucking, and port backlogs)
- Unpredictable delivery schedules increase storage time and shrink
- Weather events impact movement and handling
Potential Solutions
- Enhanced logistics planning using real-time tracking and forecasting
- On-site rail or improved truck routing to streamline flow
- Building flexible storage capacity for peak periods
- Diversifying transport modes (i.e., rail, barge, and truck)
- Stronger regional coordination among elevators and processors
4. Labor shortages and skills gaps
Challenges
- Difficulty recruiting skilled workers in rural areas
- Aging workforce and limited incoming technical talent
- High safety-risk environment requiring specialized training
Potential Solutions
- Automation and robotics for handling, sampling, and monitoring
- Apprenticeship and training programs with local colleges
- Better safety protocols and incentives to retain experienced staff
- Cross-training employees to build flexibility
- Competitive compensation and housing incentives in rural sites
5. Regulatory pressure and food safety requirements
Challenges
- Stricter requirements around traceability, mycotoxin control, and contamination
- International buyers enforcing tighter standards
- Compliance requires time, investment, and documentation
Potential Solutions
- Digital record-keeping and traceability systems (e.g., blockchain ERP, barcoding)
- Improved sampling protocols to detect contamination earlier
- Training staff in regulatory compliance and hazard analysis (e.g., HACCP)
- Upgrading equipment for cleaner handling and segregation
- Regular third-party audits to reduce risk and improve transparency
Author’s evaluation: I actually think ChatGPT did a nice job of summarizing our big challenges and some ideas to consider in terms of addressing them. We should come back to these later. Perhaps in future articles!
Prompt: “What is the allowable storage time for wheat as a function of grain temperature and moisture content?”
Result:
Good question. It turns out that there is no universal “closed-form formula” that reliably gives “allowable storage time (AST)” for wheat only as a function of grain temperature and moisture content — at least not one that applies in all storage situations. Instead, experts use empirical tables and guidelines based on past observations. That said, you can get approximate storage-time guidance from such tables, and I can explain what they look like and what they mean, and why they are limited.
Some extension and agricultural-grain publications provide approximate allowable storage times for cereal grains (including wheat) under different combinations of moisture content and temperature. These are based on historical storage experience under clean/dry/insect-free conditions.
For example — one such table (for “cereal grains”) shows something like:
“Allowable storage time” here refers to the estimated number of days before significant quality loss (mold, spoilage, heating, nutritional degradation) becomes likely — assuming ideal conditions (dry, clean, well-ventilated, insect-free, stable environment).
This kind of table gives a rough guideline: At lower moisture levels (e.g., 14-16%) and cooler grain temperatures (50-60 degrees F), wheat (or other cereal grains) may store for many months without major risk. But as moisture increases (17-19%) or temperatures rise, storage time before quality risks shorten significantly — often to a few months or less.
Why this is not a “true function” and why results vary widely
Other factors matter a lot. The tables assume “ideal storage” conditions: clean, dry, insect-free, well-ventilated. In real life, problems like insect infestation, moisture migration, condensation, poor ventilation, and temperature fluctuations can dramatically shorten storage life.
“Quality loss” is multi-dimensional. What constitutes “failure” can vary: mold growth, germination loss, mycotoxin formation, insect damage, heating, shriveling, taste change, etc. Each of these may respond differently to temperature and moisture.
Moisture content isn’t static. Grain is hygroscopic — it exchanges moisture with ambient air until it reaches equilibrium moisture content (EMC) based on ambient relative humidity. So, if ambient humidity changes, the grain moisture content might change over time, invalidating the original conditions.
Temperature fluctuations matter. Elevated temperature for even short periods (e.g., a hot day) can accelerate deterioration, even if the “average” temperature seems acceptable.
Therefore, rather than a neat function, what you get are guidelines and risk-based estimates (e.g., “If you’re storing wheat at 14-15% moisture and 50-60 degrees F, you can expect it to be safe for many months, but everything degrades quickly beyond 17-18% moisture and warmer temperatures”).
Practical recommendation: What to do if you store wheat
Dry wheat to low moisture before storage, ideally less than or equal to13-14%. Lower moisture dramatically extends safe storage time.
Store in cool, stable conditions: 50-60 degrees F is much better than 70-plus degrees F. Cooler is even better if feasible.
Ensure the storage environment is dry, ventilated, clean, and insect-free.
Monitor periodically: check grain temperature, moisture, signs of heating, mold, or insects – don’t treat by “set and forget.”
If storing for long term (many months to years), consider sealed, moisture-controlled containers (or bins with moisture-barrier features), or periodic aeration/drying if needed.
If you like, I can run a sensitivity analysis for wheat: show you approximate “safe storage lifetime” (in months) over a range of moisture (12-18%) x temperature (40-80 degrees F) assuming ideal storage conditions.
That kind of “heat map” can help you quickly see “safe/risky/likely spoilage” zones for your storage.
Do you want me to build that heat map now?
I ran the sensitivity analysis and produced an empirical model, a heatmap, a lookup table, and a CSV you can download.
What I did (brief)
Fitted a simple empirical model to representative extension/table values:
Model: AST = 200 days × 2^{s·(T_ref − T)/5} × p^{(M − M_ref)}
Anchor: 200 days at 21 degrees C (70 degrees F) and 14% moisture.
Fitted parameters from multiple table points.
Produced a heatmap of estimated Allowable Storage Time (AST, in days) over:
Temperature 4-30 degrees C (≈39-86 degrees F)
Moisture 12-19% (w.b.)
Saved a CSV of the full grid and the heatmap image.
Key results (from the fit)
Fitted parameters:
s (temperature sensitivity) ≈ 0.628 (if s=1 that would mean AST doubles for each 5 degree C decrease; fitted value <1 means temperature effect is a bit weaker than that strong rule-of-thumb for our chosen anchor data)
p (per +1% moisture multiplier) ≈ 0.742 (so AST ≈ 0.742× per each +1% moisture)
Fit error (RMSE) on the small empirical dataset: ~38.8 days (these are approximate/ballpark data points).
Example lookup table (AST in days):Important caveats (please read)
This is an empirical approximation fit to a few extension-table points — not a universal physical law. Real storage life depends strongly on other factors: insects, initial grain condition, storage cleanliness, ventilation, moisture migration, oxygen, mycotoxin risk, and temperature fluctuations.
The model assumes conditions remain stable (constant moisture and temperature). In practice, moisture will equilibrate with ambient relative humidity and temperature fluctuates, which changes AST.
Use the numbers as guidance for risk assessment (zones of “short,” “moderate,” and “long” expected storage), not as guarantees.
If you want, I can:
Produce results in months or annotate the heatmap with risk bands (e.g., <90 days, 90-365 days, >365 days).
Author’s evaluation: Okay, I am very impressed by these results. Even though corn has been the most studied in terms of Allowable Storage Time as a function of temperature, moisture content, and relative humidity, ChatGPT did some good work in terms of summarizing information, making estimates for wheat, and developing a fitted regression model to predict storage time. It was not, however, able to produce a legible legend for the graph illustrating the risk bands, no matter how many times I tried.
So What?
I know that I am a novice in terms of using AI, but most are. Universities around the world are just now beginning AI training programs. Most people have been using ChatGPT and other tools for relatively simple tasks. And ChatGPT is only one of several platforms.
Yes, it is great that we can generate images, have scientific papers summarized, generate a list of ideas, and even summarize the physics of grain drying aeration and storage, and then develop control algorithms.
But so what? How can we consolidate all known information about storage, handling, drying, and processing for various grains, and then use this information for sensing and control? How can we use AI to optimize our specific processing operations? How can we use AI to deal with a skilled labor shortage?
I think we need to dig deeper into AI capabilities to see what the various platforms, systems, and applications can do and to discuss what these might mean for our future workforce and our processing operations.
Stay tuned!
Kurt A. Rosentrater, Ph.D., is a professor in the Department of Agricultural and Biosystems Engineering at Iowa State University, Ames. He can be contacted at karosent@iastate.edu or 515-294-4019.