Logistics Challenges Receive Innovative Solution
We live in an uncertain world, and we have to adapt as information arrives: did the driver accept the load? Did the customer accept the spot price? Was a driver able to accept a load out of a new region? This gets harder when what we observe depends on the decisions we make. Perhaps the shipper would have paid a higher price, but we had to try offering them the higher price to see if they say yes. We call this "active learning."
At Optimal Dynamics, we have pioneered powerful tools for this (including an entire book), but we continue to innovate. Active learning takes on a new dimension when what we learn depends on other information such as whether we have a truck, where is the driver, and what we did yesterday. A special case of this new class of problems is something we are calling the "information-collecting vehicle routing problem.
"The first two research articles on this topic led by Optimal Dynamics' own, Chief Innovation Officer, Warren Powell and Senior Operations Research Scientist, Larry Thul explore this new class of problems and opens the conversation of how technology may solve them. If you’d like to dig deeper into these studies (warning - some serious math here), they can both be accessed at https://tinyurl.com/icvrp/.
