Waiting for the Robocars, Part II

The Advance Party

David Kerrigan
Jun 14 · 6 min read

In Part One of Waiting for the Robocars, I posed some questions about how cities and regulators might change to make the deployment of autonomous cars easier. In Part Two, I examine the autonomous vehicles that may come sooner than fully autonomous cars.

Among experts, there’s a very wide range of opinion about when driverless cars will be “ready” (leaving aside what that even means precisely), and as big a gap about how long they will take to become widespread. But alongside the prestigious race to launch the self driving car, there’s a very important “warm-up” race — to deploy and commercialise other autonomous and semi-autonomous vehicles, not necessarily designed to carry individual people.

The primary focus of these typically smaller vehicles is on the delivery of items to consumers. From food to essential supplies, several startups and established companies are unleashing myriad robots onto our roads, sidewalks and even skies as beach-head devices flying the flag for autonomy ahead of their better-known cousin, the self driving car.

Benign Bots?

It’s easy to see why delivery vehicles comprise an attractive category — there’s far more design flexibility than with cars, far less regulation and a lower threshold for consumer acceptance compared to replacing their beloved and trusted means of conveyance. A 20kg (44lbs) robot trundling along the sidewalk at a non-threatening 6kmph (4mph) is much less controversial than a 2 ton car without a human at the wheel, however poorly some humans drive.

With sidewalk bots like those from Starship and Kiwi having completed over 30,000 deliveries each, of payloads up to 10kg (22lbs), and trials underway with hundreds of the little rovers in multiple cities, these may be among the first autonomous vehicles often encountered by the public at large. Costing substantially less than most delivery vehicles, and operating at a fraction of the cost of a human-piloted car, these bots enable free, or virtually free, deliveries. Fully electric, their environmental footprint is as proportionately small as their physical size is to ICE cars typically used for deliveries. Although Kiwi and Starship are the most common providers so far, there are plenty of others in the game — FedEx have also announced experiments with a delivery bot that can even climb steps to a porch, while Amazon have recently trialled their Scout device in Washington.

Self-driving vehicles and sidewalk robots could slash last-mile delivery costs in cities by as much as 40 percent and could make up 85% of last mile deliveries.

McKinsey, 2018

But delivery bots are not without their opponents. Questions over their presence on sidewalks, their interactions with, for example, wheelchairs on a narrow pathway and vulnerability to anti social behaviour mean not everyone is welcoming them — while Virginia and Idaho have rolled out the welcome mat (or more exactly welcome legislation), there’s a ban in most areas of San Francisco.

Tentative Steps onto the Road

Larger capacity vehicles clearly don’t belong on sidewalks, but also may not need to mix with highway traffic. One interesting interim solution is the Nuro — this autonomous delivery vehicle is designed for suburban use and only about half the width of a compact car — operating on the road but at slow speeds of only 40 kmph (25 mph). Currently being tested in Texas, the R1 can carry loads up to 100kg (220 lbs) — so much more than sidewalk bots — and the success of its initial operations has seen investments in the startup of nearly $1 billion.

Taking to the Skies

Drones have quickly got themselves a tarnished reputation, with widely-reported disruptions of airports and hastily implemented ownership registration schemes to control irresponsible use. But some of the world’s largest companies are eagerly pursuing airspace solutions to deliveries. On a recent trip to Silicon Valley, I spent some time chatting to the folks at Wing, the Alphabet division developing autonomous aircraft for consumer deliveries. Already live in trials in Canberra (70,000 test flights) and shortly to commence in Helsinki, these autonomous aircraft fly at up to 120km/h (75mph) before hovering at 7m (23ft) to lower their 1.5kg (3 lbs) payload (gently) to the customer below. Just last week (June 4th), Amazon announced its near-final aircraft during the ReMars event, which they say will be ready for use within “months”.

These all-electric drones have a fairly obvious benefit for fast deliveries but will raise many privacy, safety and noise pollution concerns from people under the flight paths.

Subtle Shuttles

Away from the world of logistics and last mile delivery, Shuttles comprise another area of autonomous transport showing real early traction. Typically operating at low speeds on partially protected routes, they offer a non-threatening introduction to autonomous transport. Seating up to 10 people, they are ideal for use in large factories, campuses or even to link to suburban transit stations.

Out of Sight

There are plenty of other examples of autonomous technology slowly moving out from the labs into the outside world. But much of it is avoiding direct interactions with unpredictable humans for now, while learning to navigate simpler worlds. But if you look hard enough, you can find autonomous sweepers out early on the streets of Disneyland Shanghai, or indoors in Seattle-Tacoma airport and at some Walmart stores. Swedish startup Einride has just announced trials of a freight solution that can haul 26 tonnes.

Next Steps

So it seems we’ll see myriad other autonomous and semi autonomous vehicles before self driving cars become commonplace. As autonomy continues to advance from these simpler use cases, I believe we’ll see it expand cautiously but continuously through new use cases. I suspect it won’t be long until we see lower-risk deployments — how about night-time rebalancing (moving the cars to areas of higher demand) of shared cars such as Zipcar or RideNow? The less crowded roads with fewer vulnerable road users (cyclists/pedestrians) about would make for a more forgiving environment than day-time operations, while empty cars could always err on the side of caution without worrying about passengers.

There are fascinating changes ahead as autonomy takes hold — on roads, sidewalks and in the skies above us before we get to fully autonomous cars. In a way, it feels like these other examples of autonomous technologies are like an advance party, testing consumer reaction and paving the way for their more disruptive successors to come. What if consumers react negatively to these early encounters with autonomy? Could a backlash delay driverless cars, even if the technological and regulatory challenges are solved? Even if they’re not aware of their trailblazing role, these robots are a vital barometer of public opinion preparing us for self driving cars.

David Kerrigan

Written by

Thoughts about technology and society. Author of three books: details at https://david-kerrigan.com