Exploring the impact of autonomous technology and AI on the last-mile delivery of retail merchandise.
In “Autonomy delivers: An oncoming revolution in the movement of goods,” KPMG discusses how autonomous vehicle (AV) efficiencies will allow delivery timelines to shift in response to consumer demand for speed from next-day to same-day to, ultimately, same-hour.
KPMG strategy and retail professionals brought that discussion to life on stage at the Detroit Auto Show, exploring the rise of specialized autonomous delivery vehicles designed to reach the last mile.
The average number of shopping trips per U.S. household dropped to approximately 380 in 2018 from 480 in 2009, while online shopping now accounts for 9 percent of U.S. retail sales. The pressure to deliver all of those online orders the next day or on the same day already tests the efficiency and resources of traditional retail models, and growing demand for same-hour delivery only increases that pressure.
“The largest players are faced with the reality that if they don’t have an offering that appeals to the customers and competes effectively with newer online entrants, they’re going to lose the share,” Katherine Black said. “And at the same time, they’re faced with having to invest in all the infrastructure that it takes; delivery is an even more substantial investment.”
Today, a number of retailers try squeeze out some delivery costs by aggregating demand and bundling multiple orders. They can’t afford their own customized delivery services on such thin margins.
“It’s all in and out, there is no drop density. There is no ability to make the driver and that vehicle efficient, so what do we have to do?” Tom Mayor asked. “We have to take the driver out.”
Indeed, the driver represents the highest cost in logistics, and the solution could be a driverless, small, light vehicle that can affordably shuttle back and forth, he added. “For all of you here in the room, there’s potentially a massive market for small delivery bots that can affordably get me all kinds of products, in the time that today I get a pizza.”
Delivery logistics are only becoming more complex, according to Todd Dubner. He asked the audience to consider the number of packages delivered each day, and then given the increasing urgency of delivery, how to bundle and simplify routes. Delivery vehicles that can meet all these requirements efficiently, effectively and affordably will have to be structured differently.
“The vehicles will likely get smaller and smaller, until they are truly autonomous and... well, we think, they probably will look like a couple of bags of groceries on wheels,” he said. “The will get small enough that it’s very cheap for them to go out and back very quickly, and they will look very, very different than any delivery vehicle that’s coming to your door today.”
Dubner added that AV delivery isn’t expected any sooner than commercially viable AV people movement, approximately 2022. Indeed, China may see the spread of autonomous delivery more quickly, according to Bala Lakshman. Alibaba and JD.com already invested in and are using autonomous delivery bots.
“They are clearly ahead of the curve on this because of necessity. They have the largest e-commerce companies, they have the deep pockets, so they are actually moving further ahead than us,” he said. “It is becoming a reality, like Todd said. The question of when it will become mass market is probably a couple of years.”
Quick, autonomous delivery can get down to between 40 and 50 cents per delivery assuming Dubner’s “groceries on wheels” lightweight and inexpensive frame that makes continuous trips of five miles or less. “Those kinds of economics, if you’re asking me for a dollar to get [my delivery] in two hours versus get it tomorrow, we think drives that demand,” he said.
Still, retailers are challenged to deliver all the way from the warehouse into the customer’s hands, a harder proposition with autonomous technology. Amazon already proposed gaining access to customer homes, but other technologies are still evolving.
“These are the questions that I think are also creating big opportunities. Innovating the last couple of feet can give a competitor an advantage,” Dubner said. “That's where a true system provider that helps coordinate from the fill of the delivery vehicle to the empty of the delivery vehicle, and the coordination of the logistics from point A to point B, may have an advantage.”
At approximately $2,000 apiece, delivery bots today are “extremely cheap,” Lakshman said. Because there are no requirements to carry or protect a human, these simple machines require only a lightweight electric motor and a guidance system, with an operating cost of approximately 5–10 cents per mile.
However, “The jury is still out as to whether it’s going to be an OEM’s play or a new entrant,” he added. Auto manufacturers have to guard themselves against trying to over-engineer for long-lasting vehicles that can handle significant mileage and weather, while upstarts produce a vehicle that may be suitable for limited geographies and have shorter lifespans but are so inexpensive they have a stronger appeal.
Black added that the latter makes sense in the consumer-driven retail industry which, by its nature, requires customization. “You can design sort of a one-size-fits-all [vehicle], which is probably where the market is today. But I would imagine that ultimately that’s going to need to be much more specialized with quite a bit of variety in it, and much tighter integration with the [retail] business models.”
How demand for these autonomous delivery vehicles will impact the automotive supply chain still needs study, the panelists said. “It’s not going to be the traditional auto model where you sell the vehicle and you’re done with the vehicle… and you sell parts and services,” Lakshman said. “It’s going to be a total cost of ownership game.”
To access this panel presentation, please click here.
To access the KPMG whitepaper on this topic, Autonomy delivers, click here.