A Boston Startup is Solving a Classic Optimization Problem with Machine Learning
Here’s a classic optimisation problem – most transportation and logistics companies are still using the old methods of deciding routes for trucks. While the big companies are getting on board with data analytics, others are stuck in a rut.
When drivers get stuck in traffic jams, or the original path has a diversion in the middle, the truck drivers get behind on their schedule. This, as you might expect, results in major losses for the company and leads to customer dissatisfaction.
A startup from Boston, Wise Systems, has devised a solution. They have paired data (partly taken from the drivers’ mobiles) with machine learning algorithms to optimise each route every truck will take. The variables the ML model looks at include the driver’s speed, GPS location from the cell phone, traffic on the route, weather, the truck’s destination, and what time the customer will be available to receive their order.
So the question is, how is this any different from the old methods? Well, the biggest advantage is that the model allows flexibility. The routes can be changed immediately depending on any changes in any variable. For example, if any road is closed or diverted, the driver will receive an immediate alert on his phone. If the driver is not scheduled to deliver on time, he will get an alert asking to speed up his truck.
This stems from the age-old Traveling Salesman problem which a lot of experts have been trying to solve since the early 20th century. The possibilities are endless.
Using Wise System’s ML model, companies can save a lot of money by minimising the delivery time (and as a result making more deliveries in a day).
Our take on this
As mentioned, this is a problem that has plagued many a company since decades. It’s a surprise that it’s taken this long for a proper solution to come up for medium sized companies (the big companies like Amazon and Kuehne Nagel) have their own models). But now that it has, we can expect it to catch on pretty quickly.