Inspiration
KITT is inspired to actually disrupt the existing way of big online retailers, shipping companies and trucking companies are optimizing loads in their fright fleet and containers.
What it does
KITT uses the dreamer v3 physics base world model to predict the most optimal way of loading a container, a truck with all your shipment data and also uses weather prediction to adjust the load depending on the route and predict traffic as well, the goal is simple optimize loading in trucks to the point you dont have any kind of loading related delays or damages to the item at any cost.
How we built i
We built it using and leveraging opensource models that are both LLM and also non diffusion world models, our physics engine can predict the most optimal layout and loading weight and items for any available truck in your fleet by calculating every single metrics from your files the wifi at the venue was so hard to work with
Accomplishments that we're proud of
What we learned
So far world models are not mainstream unless they are diffusion based which raises the question why
What's next for KITT
KITT can power the future of fright loading and make it completely autonomous, an ai engine that can predict the best way to load a truck or a container and then passing it to robots that can actually load it and also an autonomous truck that can deliver that load is surely something we will step to.
PLEASE WATCH OUR VIDEO DEMO @ https://www.loom.com/share/c094b371b80c4a3fb29b07197b8fc2d5
Log in or sign up for Devpost to join the conversation.