Valtra, based in Finland, is a pioneer in developing easy-to-use and tailored tractor solutions. Valtra is part of the AGCO Corporation, the world's third largest designer, manufacturer, and distributor of agricultural machinery.
Our collaboration with Valtra started with the ValtraHackathon, where the challenge was to create added value from the telemetry data generated by the machines. Specifically, we were looking for solutions that would increase profitability and contribute to sustainable development.
Loihde won the hackathon with its idea for a SmartCoach app. The app guides and encourages the driver to operate the tractor better. Loihde’s idea convinced the jury with the following criteria.
“Loihde’s winning idea uses data from Valtra Connect, solving a customer's problem so that the tractor is used more efficiently, resulting in savings for the environment. The solution also supports Valtra's Best Tractor Experience vision," says Aapo Aijasaho, Manager, Valtra Product Marketing.
Loihde and Valtra got down to business and, during 2022, implemented a pilot project based on the winning idea. The project consisted of three parts:
The project team explored what could be gained from the data that would be interesting and useful from the perspective of the tractor driver or owner. We identified use cases, such as monitoring idling time and assisting in idling reduction, gear use and analysing speed variations. The team initially selected idling as the most promising “low-hanging fruit” and decided to focus on its analysis.
The team was tasked to answer questions like: What do end users want to see and how are things presented? What does a potential app look and feel like?
The project team used design tools to develop several examples of views and app functionality. The examples were developed following Valtra's brand guidelines.
Idling was selected for data analysis because it was considered to have significant potential from an economic and environmental point of view.
The team created an algorithm to identify whether a machine is idling based on the data generated by the machine. The algorithm also classifies idling into different categories: necessary, suspicious, or unnecessary idling.
The algorithm was applied to a large sample of tractor telemetry data, allowing an assessment of how much room for improvement there is in terms of idling. The analysis was able to identify a subset of tractors that could benefit significantly from a reduction in idling.
The SmartCoach project allowed Loihde to make extensive use of its expertise. Successful collaboration required specific expertise in conceptualization, application design, data processing and analysis, and visualisation of results. The technologies used in the project include Figma, Python, AWS and PowerBI.
The result: