Inspiration

In an American Society of Mechanical Engineers article about farm tools repair a farmer could spend close to $100,000 per year on repair costs if they have high tech tractors on their farms. There is an increasing demand for built-in software diagnostic tools that could potentially half the cost of repair to farmers. There is also a growing concern about farm cybersecurity. According to a USDA report on threats to Precision Agriculture, hundreds of farmers purchase diagnostic software on pirated software markets to jailbreak their tractors and sprayers and conduct diagnostic work on-site instead of taking their equipment to a dealership for service. This makes equipment vulnerable to malicious actors during planting and harvesting season

What it does

The app takes in 8 features (Engine Load, Engine Coolant Temperature, Hours engine is operational, Gage pressure of oil in the engine lubrication system, Fuel Delivery Pressure, Hydrostatic pressure, Temperature of hydraulic fluid and SCR_TANK_LEVEL) and predicts whether the tractor is healthy or not at a particular point of time

How I built it

We used Jupyter Notebooks to build our model and React to build the app

Challenges I ran into

Modifying the states in React and processing the data and getting relevant features

Accomplishments that I'm proud of

Built our first end to end product and learned firsthand how machine learning can be integrated into webapps

What I learned

First Hackathon we participated in so we learnt how to build an end to end product for the AgTech industry.

What's next for DiAgnostics

We are excited to present and look at what other groups have done

Share this project:

Updates