Geospatial Mapping & Capacity Estimation of On-Farm Grain Storage in the US

We are looking for a GIS, remote sensing, or computer vision specialist/team to help identify and map on-farm grain storage infrastructure in the United States. The goal is to detect grain bins and other on-farm storage assets from aerial or satellite imagery, estimate storage capacity where feasible, and produce farm-level lead data for farmers with likely corn, wheat, or rice storage capacity. We are especially interested in starting with Kansas, North Dakota, and Arkansas, but are open to recommendations on the best pilot geography. Project Objectives We want to build a geospatial dataset that includes: Farm-level locations with likely grain storage Detected grain bins or storage structures Estimated number of bins per farm/site Estimated bin diameter, footprint, and capacity where feasible Estimated total storage capacity per farm/site Crop/storage relevance, especially corn, wheat, and rice Confidence score for each detected site Output usable for lead generation and sales outreach Pilot / Test Project We are very open to starting with a paid test project before expanding the scope. For the test, we could provide a farmer contact list to help narrow the search area, validate results, or match detected storage sites to known farm operators. We would like the contractor to advise whether starting from a contact list, parcel data, county-level imagery, or another approach would produce the best results. Potential pilot areas could include: Kansas wheat/corn storage regions North Dakota wheat/corn/soy regions Arkansas rice storage regions A small set of counties in one or more of these states The test should help determine accuracy, cost per mapped farm, feasibility of capacity estimates, and the best path for scaling. Possible Methodology We are open to different approaches, including: High-resolution aerial imagery analysis NAIP or other public imagery sources Commercial satellite or aerial imagery if needed Computer vision/object detection models GIS-based structure detection Manual or assisted QA Matching detected storage sites to farmer contact lists, parcels, or farm addresses Estimating storage capacity from bin footprint, diameter, height assumptions, or standard grain bin references Deliverables For Test Project Farm-level list of likely grain storage leads GIS file of detected storage sites CSV with farm/site location, detected bins, estimated capacity, crop relevance, and confidence score Match-back to farmer contact list where feasible Notes on false positives/false negatives and expected accuracy Recommended process for scaling across KS, ND, AR, and other US grain regions Estimated cost and timeline for larger rollout Ideal Contractor Experience GIS and geospatial data processing Remote sensing or aerial imagery analysis Computer vision/object detection from aerial/satellite imagery Agricultural infrastructure mapping Farm parcel, rural property, or crop-region analysis Python geospatial stack, QGIS, ArcGIS, Google Earth Engine, PostGIS, etc. Experience creating lead datasets from geospatial signals Proposal Questions How would you approach mapping on-farm grain storage at the farm level? Would a farmer contact list help, and how would you use it? What imagery or data sources would you recommend? Can you estimate storage capacity, not just identify bins? If so, how? What pilot geography would you recommend across KS, ND, and AR? What accuracy level do you think is realistic? What would a paid test cost and how long would it take? Have you done similar agricultural, GIS, or object-detection work? Success Criteria A successful test would produce a usable farm-level lead list showing likely on-farm grain storage, with defensible confidence scores and capacity estimates, plus a clear recommendation on how to scale the work across broader US grain-producing regions.