Seattle

Eno Transportation Weekly

How Can Automation Improve Freight Delivery?

At Eno’s Capital Convergence conference, held last month, one of the sessions examined the impact that existing and imminent new technologies will have on the freight sector.

Panelists were:

  • Steve Boyd, founder and Vice President of Peloton Technology.
  • David Catania, former District of Columbia Council member and public outreach lead (US) for Starship Technologies.
  • Chase Murray, Assistant Professor of Industrial and Systems Engineering at the University of Buffalo.

The panel was moderated by Geoffrey Milsom, Director of Transportation Solutions at enVista.

Given the ground-based technologies of both Peloton and Starship, most of the discussion focused on ground shipments, but since Peloton’s model focuses on interstate commerce (helping tractor-trailers to drive in tandem more efficiently, at least as an initial phase of development) and Starship focuses on the last mile or less (delivering packages in an urban setting), the two technologies are aimed at very different kinds of purchasers (and face completely different regulatory and legal schemes).

Peloton is focusing on companies that own major fleets of long-haul trucks (but eventually they hope to sell to owner-operators as well). Their technology uses vehicle-to-vehicle (V2V) communications in cooperation with a kind of adaptive cruise control to allow trucks to travel as closely together as the response times for pneumatic brakes allow. The short-but-constant distance between trucks allows significant fuel savings (two trucks “platooning” in tandem collectively use 7 percent less fuel than they would driving independently, and Boyd pointed out that the trucking industry spends about $100 billion on fuel each year).

And electronically linking the braking systems in every truck in the platoon (which are also linked with other state-of-the-art collision avoidance systems that are a prerequisite for installing Peloton tech in a truck) allows each truck after the front one to brake significantly earlier than they would if they had to wait for the driver’s eye to see the brake lights ahead, make the right foot move and slam on the brake pedal, which increases safety. If the safety benefits can be conclusively demonstrated, trucking companies might save on liability insurance premiums if they install the technology.

Peloton’s technology is focused on trucks weighing tens of thousands of pounds driving 65 miles an hour over long distances, but Starship’s tech features robots weighing 30 pounds traveling less than 4 miles an hour over distances of less than a mile on public sidewalks. Each small, battery-powered, six-wheeled drone can hold about three bags of groceries (or, from the looks of it, a case of long-necked beer bottles). When reaching its destination (guided by hyper-accurate mapping of the service area that will be significantly more accurate than GPS), the recipient presses a button on an app on their phone to pop the hatch on the drone (or the drone operator releases the hatch remotely to a building concierge). (ETW’s Greg Rogers profiled Starship’s delivery technology last October, and one of their drones was rolling around the conference room.)

While Starship’s DC pilot program has a delivery area that is limited to the downtown core of the District, which is predominantly non-residential, Catania pointed out that their business model assumes a lot of personal deliveries in addition to business purchases, because many people, fearing “porch pirates,” have their packages delivered to their office instead of their home.

In the initial phase, Starship is marketing its DC (and Redwood City, CA) pilot programs to restaurants and grocery stores. (Catania said that Giant’s competing grocery delivery service, Peapod, has a delivery cost of between $8 and $13 per delivery whereas Starship hopes to get its delivery cost to around $1.) He said the eventual goal is to break into more generalized package delivery – he noted that the growing e-commerce field will eventually have a “rendezvous with physics” in that there is only so much space for delivery trucks in downtown urban areas, and labor represents 61$ of UPS’s costs and 40% of FedEx’s costs. However, they intend to operate Starship as a service where they run their own centralized dispatch and control centers (along with fleets of converted Mercedes-Benz Sprinter vans that carry multiple delivery drones and release them in neighborhoods).

He also noted that two of the biggest problems with e-commerce deliveries are missed deliveries and returns, and their technology deals with both issues cheaply (ideally, customers summon the drone with their phone when they are ready and then wait for it, and they get to see the cargo before the drone returns to base).

Boyd and Catania described two different regulatory agendas for their technologies. He described Peloton’s system as a combination of several SAE Level 1 and 2 technologies that the federal government, and many states, already allow under current law since the driver is still responsible for hitting the brakes or turning the wheel if he/she sees a danger that the automated system has missed. However, he said that the V2V portion of the system must meet federal standards. (And he said the big leap is between SAE level 2, where the driver is still constantly engaged, and level 3 where the driver becomes disengaged – if Peloton technology evolves to level 3, significant legal and regulatory changes will be needed). 11 states have passed legislation on truck platooning, with Michigan’s being the most comprehensive. Peloton is working closely with Michigan, and neighboring states of Ohio and Pennsylvania, as they develop their “Smart Belt Coalition” to share information on pilots around the testing of platooning.

Catania said that Starship’s drones are too small and too slow to be considered a “motor vehicle” under federal law, so federal regulation is not a problem. But he said that many states don’t have that de minimis threshold so Starship has to work with state governments on a case-by-case basis to get their drones exempted from motor vehicle standards there as well – similar to how Segway went state-to-state getting such exemptions. (He said they had had a good vote in committee in the Virginia Senate on the day before the January 26 conference.) They are also beginning to work with states of Florida and Wisconsin on legislation and piloting drones there.

It was left largely to Professor Murray to discuss the situation with aviation delivery drones. While highlighting the potential public health benefits, he did state that that the few public trials of delivery by drone were mostly for show. UPS delivering blood to a clinic in Rwanda, a delivery of insulin to some small island off the New Jersey coast, a delivery by Amazon Prime Air to a location conveniently next to their warehouse. All shared a fixed takeoff point, a well-scouted destination, and no obstacles in between – and in the U.S., they were able to stay within the line of sight of the drone operator, a critical factor under the existing FAA regulations. Murray indicated that the current FAA rule requiring a drone to remain within the direct line of sight of the operator at all times would pose an insurmountable obstacle to many of the kinds of delivery services that were originally envisioned – but that logistics companies were exploring other options (Wal-Mart, for example, is using drones within its own cavernous warehouses).