project · 2023 · Solution Architect · Fraunhofer IESE / RPTU Kaiserslautern
Smart Sensors in the Field: Soil Data Validation Capstone
Solution architect on a Fraunhofer IESE capstone in partnership with John Deere and RPTU Kaiserslautern, building a pipeline to parse soil lab reports, correlate them with vehicle sensor data, and validate smart sensor accuracy through Power BI dashboards.
Precision agriculture depends on trusting the sensors on the machine. This capstone tackled that trust question directly: do the readings from field sensors on John Deere vehicles actually match what a soil lab measures?
The project was a collaboration between RPTU Kaiserslautern and Fraunhofer IESE, with Fraunhofer working in partnership with John Deere on the industry side. The role on this project was solution architect, responsible for the system design and technical direction of the data pipeline.
System architecture
The system is built around a digital sampling platform with four components. A mobile application provides the user-facing interface for creating sample data, uploading PDF lab reports, and viewing analytical results. A Node.js service API framework handles sample data submission and PDF upload, persisting everything to a database. An analytics dashboard sits alongside the mobile app, consuming the stored data and surfacing analysis results back to the user.
Parse, correlate, validate
The core system parses soil lab reports from PDF documents and correlates the extracted measurements against sensor data recorded from field vehicles. Bringing two very different data sources into a comparable representation required careful schema design and field-level alignment between lab format and sensor output format.
From data to insight
Power BI dashboards and data analysis pipelines visualize how field sensor readings measure up against lab benchmarks across different soil conditions and locations. The output is a quantified view of sensor accuracy that makes calibration gaps visible and gives field teams an actionable basis for adjusting or trusting their sensor readings.