Inspiration
We really liked the CustDev part with the mentors. It was not only useful, but very friendly and nice.
What it does
We predict the time the truck is going to be driving from the plant to the construction site. We also have added the opportunity for the plant manager to choose maybe not the fastest, but the most convenient times for the customer.
How we built it
We trained ML-model and created the web app prototype for one plant.
Challenges we ran into
The data we were provided with was small and did not cover the COVID-19 situation. Also there were poor internet connection and cold nights.
Accomplishments that we're proud of
We have MVP and our min absolute error is only 9 minutes!
What we learned
Concrete is a very perishable product.
What's next for Truck and track
Truck and track is only for concrete, but the company also provides other construction materials. Using the same methodology that we used now, we can widen the range of "perfect time of delivery" products for customers. We also can use more complex algorithms to minimize the weigh of human communication in the company.
Log in or sign up for Devpost to join the conversation.