Daniel Taylor, product manager-routing with Ortec, discusses a new approach to the age-old practice of time-slot booking — both at the e-commerce customer's door and the warehouse dock.

SCB: How can companies use the concept of time-slot booking in order to increase the efficiency of deliveries in the supply chain?

Taylor: There are a couple of places where it can be applied, mainly in order fulfillment. You really want to get it right the first time to minimize your cost. You see that mainly in retail to home delivery and the services industry — for example, in pest control and home healthcare. But it also goes beyond that, to the booking of docks. If a carrier has to deliver or pick up goods, the same system can be used to say, "OK, which docks are available so that I can plan that I'm there at that time, and don't have to wait at the gate until I can go to the dock?"

SCB: This is something that's been a practice for a long time, but the experience hasn’t always been satisfying to the receiver. Especially in the case of home deliveries, the time slot might be very wide. Which is not acceptable these days to a receiver of goods.

Taylor: What you typically see is that companies will offer a very broad time slot, say, eight a.m. to four p.m. Only if the customer asks for a narrower window will they cut it down. But now, with developments in technology and artificial intelligence, you can very quickly determine optimal time slots, taking into account order characteristics, geography and routing.

SCB: How narrow can that window be when you're scheduling a particular delivery time?

Taylor: It can be really narrow if you want it to be. It's just that the risk of not being able to fulfill the order increases. Depending on the type of service you deliver, you can get down to half an hour or an hour.

SCB: That's tighter than most consumers are used to receiving in terms of home delivery.

Taylor: But it is possible.

SCB: So it is coming down to same-day, or even a particular hour?

Taylor: Yes. And it's all made possible through developments in AI. For example, you can use AI techniques to do the profiling, determine what order sizes you have, and what time slots people want. Once you know that, you can adjust your capacity and resources, and deploy them correctly so that you can fulfill what you're promising.

SCB: This could be of great value at the receiving dock, where truckers are often required to wait for long periods of time when a space isn’t open. It affects the carrier's willingness to serve the shipper in the long term.

Taylor: Yes, for sure. There’s no reason why the same techniques can’t be applied there, where profiling is much simpler. And it goes a long way toward enhancing the relationship with your carrier, who often feels that it’s beholden to whenever you’re available. If you can show a schedule of a whole day, with all the time slots that are available, it feels like they have a choice. That's the same for consumers as well. You can also control which time slots you actually want to show, based on your business rules and the ones that work best for you.

SCB: An important aspect of AI today is machine learning. Does that imply that the system can in fact learn from experience, and get better at coming up with the right output for scheduling delivery times?

Taylor: Absolutely. The way we mainly use it now is for learning, so based on the profiling I was mentioning before, you can adjust your capacity. But the algorithm itself can also improve, based on the information it receives. And if you start coupling that with revenue management, which prices time slots for delivery to consumers, that can be very powerful stuff.