'Safety will always come first,' insist Arizona biz org in response to Uber self-driving car death

Robot rides slow down but keep moving

Intel Capital "My greatest fear is that we will hit the winter of AV (autonomous vehicles)," says Jill Sciarappo in response to an unwelcome question at the Intel Capital conference in Arizona – the US state that saw the first self-driving car death last year.

Sciarappo is the marketing director of Intel-owned Mobileye, one of the main companies in the autonomous vehicle market, and she was sat alongside the CEO of trade organization the Arizona Commerce Authority (ACA) Sandra Watson, and professor from Dr Sethuraman Panchanathan from the Arizona State University (ASU), at the Biltmore Hotel in Arizona just a few miles from the fateful crash.

Together they are representing the new "Institute for Automated Mobility" which was set up in October last year and represents Arizona's response to the death of Elaine Herzberg, although no one is willing to say so publicly.

The institute combines business with government and academia but Watson – a government figure who oversees it – makes no bones about the fact that it is there to help drive the self-driving car industry. It was "set up to accommodate industry as much as possible," she noted and repeatedly referred to the Arizona governor's business background and desire to use his position to help corporates.

Just in case you're in any doubt as to its approach, the main webpage for the institute notes at the top that "94 per cent of serious vehicle crashes are due to human error."

All of which is important and relevant because Arizona remains the main testing ground for autonomous vehicles in the US and by extension the world. That's down to three factors: first, Arizona is flat, its roads are straight, and new, and well marked, and the climate is dry and sunny. It's the perfect test-bed for getting machines to learn how to drive cars.

Second, Arizona's governor has championed the industry and gone out of his way to make Arizona a good place for AV companies to set up shop, even pushing and passing legislation to accommodate the industry. And third, the US is the center of technological advances and research on autonomous cars; it's where the money and the engineers are.

Don't say nuffin

Which is why no one wants to talk about Elaine Herzberg and her death at Uber's hands. After a lengthy presentation about what the new institute will do and how it will do it (short version: test beds, tech specs, industry best practices), no one had mentioned the death that still casts a shadow over the self-driving dream, leaving it this reporter to raise the issue and make everyone uncomfortable.

The governor had gone out of his way to publicly invite Uber to Arizona after the company's self-driving program was kicked out of San Francisco for failing to even apply for licenses. Uber came, and then it ran down a citizen and killed her, and the governor was forced to disinvite the company and pray that anger over the issue wouldn't become a political problem for him.

So we ask what lessons have been learnt from the death a year earlier. Watson instantly glances at her notes and informs us that the Arizona state government's "focus was always on safety," and that she can't comment further because the state is still involved in litigation over the matter.

Academic Dr Panchanathan pivots to talk about how Arizona is a great place for testing autonomous driving thanks to having "lots of sunshine but also snow as well as mountains" along with a "focus on safety." And Sciarappo – as the industry voice on the panel – notes that "it is incumbent on all of us to always be safe."

But, to Sciarappo's credit, she does address the issue, albeit in a roundabout way. She refers to artificial intelligence and how it was the big exciting industry not so long ago but after it didn't meet initial expectations, interest plummeted and the industry entered the "winter of AI."

Issues around safety and autonomous cars – especially when a pedestrian is mown down during testing – pose the same risk and has led to her fear that her industry will also enter its own "winter."

It's clear that the new institute is supposed to insulate the AV industry – and the governor – from possible blowback but such is the sensitivity that no one will talk about it, or Uber, or Ms Herzberg. Which is a shame.

Revving up

But while the self-driving industry may have taken its foot off the accelerator – with a new self-driving taxi service in Phoenix slowly and carefully launching - it has most definitely not stopped moving and it may soon speed up again.

Earlier in the day we spoke in some depth with another AV company, Aeye, which designs the self-driving systems that combine radar and cameras to give cars their own machine visibility and intelligence and so help move toward full autonomy.

Last year at the same event Aeye was confidently predicting that everyone in the industry would soon move to an integrated system of both cameras and lidar in order to make progress. At the time, some were arguing that only cameras were needed, and that there didn't need to be such a strong focus on an overall system.

Aeye's prediction has turned out to be right – and arguments in the other direction may have led to the Uber death for which it was eventually found not to be criminally negligent - which makes its predictions this year all the more relevant. Blair LaCorte from Aeye argues that the past year of the industry has been messy and scrappy but that it has started settling on a basic structure.

And that structure will be a market split in two: on the one side, companies that own the "assets" i.e. cars with autonomous car technology built in, and which aim to make money from them. On the other you'll have service companies that will build and improve the AV technology itself. The days of one company doing both are over, he predicts.

He also predicts that the tech itself will largely split: between self-driving improvements or add-ons to existing cars (ADAS or Adapted Drive Systems in the lingo) that will do things like help people park or warn them if they are straying out of a lane; and fully autonomous cars that will be owned and run by large companies for things like taxi services. In other words, don't expect to go buy your own self-driving car anytime soon.

Getting better all the time

The rules of the road are becoming increasingly clear: Aeye's LaCorte notes for example that every car ad features a graphic in which a pulse comes out of the car, representing its increased safety and awareness of its surroundings. Car companies have decided that safety is going to sell. And that has given other companies more time to focus on the technology – which has led to some significant advances.

Aeye has a pretty impressive demo where it shows you what, in real time, its autonomous car is seeing. As you stand by the vehicle, you can see your digital likeness appear on a screen next to it – it sees you. If you move it tracks you. If you do something it's worried about, it flags you up as a possible problem.

The systems for figuring out how to make machines drive better are maturing and improving in large part because they have started better mirroring how we as humans drive: by constantly scanning the environment around us and paying more attention to anything that appears unexpectedly, especially if it is moving toward us at a fast pace.

LaCorte also notes that 40 per cent of what we as humans "see" is driven by other senses. If you smell coffee, you look around for coffee; if you hear something, you look toward it and assess it. Color is another factor: if you see a red light, you note it and prepare to adjust your behavior and if you see a green light you do similarly, but in a different way.

We may not see cars with synthetic noses any time soon but microphones? It's very possible. Anything that tips the balance when it comes to safety. Aeye's system has also started looking for, and making sense of color, in order to give it an extra sense.


FYI: You could make Tesla's Autopilot swerve into traffic with a few stickers on the road


But cars loaded up with new sensory system may be just the beginning of a bigger shift. In the same way that the internet was a system designed to send information from point to point but ended up being a platform on which all kinds of communications and interactions have been built, so LaCorte feels that when we do have vast fleets of fully self-driving cars on our roads we could see a similar shift.

If that sounds a little far out there, he also notes the investment put into self-driving cars could have significant changes to movement and transport. If you own a fleet of such cars that are pretty much able to go do their own thing, the current approach when the vast majority of cars are sat immobile most of the day starts becoming an enormous waste of resources.

"How would we all change our habits if travel was readily available and almost free?" he ponders.

Meanwhile, back at the Institute for Automated Mobility, the industry-led consortium hopes that its efforts will enable it to get past this uncertain point in AV evolution and move it to a place where cars driven by machines are viewed as something that could actually improve road safety and traffic congestion rather than pose a threat to their lives. ®

Broader topics

Narrower topics

Other stories you might like

  • A peek into Gigabyte's GPU Arm for AI, HPC shops
    High-performance platform choices are going beyond the ubiquitous x86 standard

    Arm-based servers continue to gain momentum with Gigabyte Technology introducing a system based on Ampere's Altra processors paired with Nvidia A100 GPUs, aimed at demanding workloads such as AI training and high-performance compute (HPC) applications.

    The G492-PD0 runs either an Ampere Altra or Altra Max processor, the latter delivering 128 64-bit cores that are compatible with the Armv8.2 architecture.

    It supports 16 DDR4 DIMM slots, which would be enough space for up to 4TB of memory if all slots were filled with 256GB memory modules. The chassis also has space for no fewer than eight Nvidia A100 GPUs, which would make for a costly but very powerful system for those workloads that benefit from GPU acceleration.

    Continue reading
  • GitLab version 15 goes big on visibility and observability
    GitOps fans can take a spin on the free tier for pull-based deployment

    One-stop DevOps shop GitLab has announced version 15 of its platform, hot on the heels of pull-based GitOps turning up on the platform's free tier.

    Version 15.0 marks the arrival of GitLab's next major iteration and attention this time around has turned to visibility and observability – hardly surprising considering the acquisition of OpsTrace as 2021 drew to a close, as well as workflow automation, security and compliance.

    GitLab puts out monthly releases –  hitting 15.1 on June 22 –  and we spoke to the company's senior director of Product, Kenny Johnston, at the recent Kubecon EU event, about what will be added to version 15 as time goes by. During a chat with the company's senior director of Product, Kenny Johnston, at the recent Kubecon EU event, The Register was told that this was more where dollars were being invested into the product.

    Continue reading
  • To multicloud, or not: Former PayPal head engineer weighs in
    Not everyone needs it, but those who do need to consider 3 things, says Asim Razzaq

    The push is on to get every enterprise thinking they're missing out on the next big thing if they don't adopt a multicloud strategy.

    That shove in the multicloud direction appears to be working. More than 75 percent of businesses are now using multiple cloud providers, according to Gartner. That includes some big companies, like Boeing, which recently chose to spread its bets across AWS, Google Cloud and Azure as it continues to eliminate old legacy systems. 

    There are plenty of reasons to choose to go with multiple cloud providers, but Asim Razzaq, CEO and founder at cloud cost management company Yotascale, told The Register that choosing whether or not to invest in a multicloud architecture all comes down to three things: How many different compute needs a business has, budget, and the need for redundancy. 

    Continue reading

Biting the hand that feeds IT © 1998–2022