Inspiration
To best meet the needs of real truck drivers, our solution to the 123Loadboard challenge focused on ensuring that the drivers get notified of jobs that truly maximizes the value of their workday. We incorporated heuristics and design choices inspired by the various trucker lifestyles.
What it does
The program, accompanied by a web app, creates notifications each containing a specific truck and a specific load. The truck is only included in notifications that is high-value to the individual driver. The program selects jobs that match the truck's specifications, the truck driver's preference on long-/short-distance jobs, as well as jobs that maximizes hourly profit or has high chance that the truck will be able to find a new job at that job's destination.
How we built it
The program was built using Python, javascript and node.js.
Challenges we ran into
The convolution of the different motivations held by drivers presented a complex problem which required us to design a robust heuristic. Furthermore, as both members of our group were unfamiliar with front-end development, making a presentable product took up a considerable portion of our time.
Accomplishments that we're proud of
Our heuristic contain consideration for preferences that go beyond what is obvious from the data flow at any given time. In particular, there is a consideration of distance from a likely "home-position" specifically used for short-haulers, as well as the use of past data to direct truck drivers toward areas that is likely to have a high density of loads around the time of arrival.
What we learned
The project was an opportunity to really get to know leafletjs and node.js, as well as get our hands dirty with some front-end shenanigans.
What's next for TruckTrackMatch
With over a hundred million trucks registered in the United States, it is clear that this is a problem for which our program needs to be highly optimized to handle large numbers.
Log in or sign up for Devpost to join the conversation.