The top two guys at FreshStart Logistics, Allen and Bill Ho, came to visit me. Their company sells an optimization engine for TMS.
I have a geeky fascination with optimization. I’m don’t have a Ph.D in Operations Research, but I’ve seen the way a number of TMS solutions approach optimization. There were several things I liked about FreshStart’s real-time engine, including the way it can use GPS and traffic data for dynamic routing. But the thing I appreciated the most about our meeting was our conversation about the last mile omni-channel logistics delivery problem.
Bill and Allen’s last company provided a dynamic routing solution for paratransit. Paratransit is passenger transportation that does not follow fixed routes. In the U.S., the Rehabilitation Act of 1973 said that public transit systems, if they received federal assistance, needed to be handicapped accessible. Most cities opted for a flexible system of small paratransit vehicles where users could schedule pickups, often in a half-hour pickup window. In big cities, this becomes a massive routing problem.
One of the capabilities of their previous company’s solution, which their current optimization engine still supports, is the ability to preposition vehicles for the next day’s routes. The way this would work in practice is that at the end of the day, a driver would drop the vehicle at an optimally located city-owned lot, then another paratransit vehicle would come by and pick the driver up and take him back to the terminal where his car was parked.
Let’s take that capability and think about last mile omni-channel logistics. Omni-channel is an initiative by brick and mortar retailers to better integrate their stores and e-commerce channels. So a retailer might support buy online, pick-up at store; or order online, deliver to home from a store; several other fulfillment paths are also possible.
Amazon has been stealing market share from traditional retailers for years. Amazon is experimenting with same-day shipping, but mostly it offers two-day and longer shipping. In an article published earlier this year I pointed out that large traditional retailers, with all their stores in metropolitan areas, could potentially compete on speedy fulfillment. Stores can be thought of as forward inventory deployment locations. Retailers could move to not just same-day deliveries, but potentially deliveries within 90 minutes of an order being placed. Or customers might sign up for deliveries during the day during a specified 30-minute delivery window.
When I had thought about deliveries from stores, I had visualized it as a delivery person leaving a store, making his deliveries, and then returning to the home base store. But it strikes me that the paratransit model could be much better. Delivery people would interleave moves between different stores and customers based upon the outputs from a dynamic routing engine. At the end of their shift, they would leave the store delivery vehicles in the most optimal store locations based upon known next day orders.
Just as imitation is the best form of flattery, so is stealing ideas from a different industry and applying them to yours.
Another interesting twist would be to actually have drivers meet during their routes and exchange vehicles so they could “return” to a location closer to their base station. Real time seems to give the impression that even during the route, drivers could be directed to pickup other freight along the way that has become available while the routes are in progress. That would be a very big advantage to the system.
I like the FreshStart routing concept, and the Paratransit kind of flavor. How about this — Instead of the store chain (Krogers, Sears, Best Buy, etc.) creating their own fleet or using students and neighbor semi-volunteers to make deliveries, let’s introduce a 3rd party we will call “Local Delivery”.
Local Delivery uses FreshStart’s routing and has a fleet of small (Ford Transit Connect type or smaller) vehicles, preferably electric or hybrid, to make deliveries. Figure in battery exchange or recharge stations.
Retailers would submit delivery requests (tenders) to a central point via a network connection or portal. The System would then determine the closest unit, most practical route, etc. and notify the driver via dash mounted display or smart phone.
The driver would pick up, ack the system, be tracked via GPS, make the delivery, ack the system, get instructions for next drop. Could even insert some interleaving of return pickups from customers and deliver to retail locations.
FedEx, UPS, OnTrac may be good candidates for this operation, however they use larger less efficient trucks to deliver a large set of orders from a central point to doors. Similar but different concept.