There’s much to be done before we have autonomous cars on public roads, ranging from testing and validation to developing business models, writing regulation and defining liability. But enough with the problems, say the AV proponents – let’s talk about solutions! Delegates at Autonomous Car Detroit heard about the challenges, and the opportunities.
Autonomous cars save lives
When it comes to autonomous cars, safety is the number one issue, said Huai Peng of Michigan’s Mcity – and no-one at the event disagreed. There’s still plenty to do to perfect not only the technology but also the regulatory frameworks. Stephen Lesh, Head of Vehicle Programs at Uber’s Advanced Technology Group, who gave his keynote at the event just days before tragedy struck in Arizona, noted that, “There are no definitive guidelines on how to prove out the safety of a self-driving vehicle. We need to engage in a proactive way with governments, cities, regulators and academics to build out the regulations on both the safety and business side.”
Better quality time spent on the road; increased productivity; increased mobility for the elderly and the infirm, for the blind and the blind-drunk – autonomous cars that drive better than humans will enhance the transportation experience and begin to cut into that 94% statistic so often cited as the percentage of human error-related car crash deaths. “The main focus of autonomy should be to save lives,” said Volker Sasse of NavInfo and the Open AutoDrive Forum. Get the technology right, said Sasse, and “the risk of human error can be reduced and will save many lives.”
Human factors matter
Develop a great business model, develop a great vehicle, and perfect the autonomous drive technology – but if you fail to factor in the human experience, then your vehicle is as good as redundant, and your business model won’t meet its revenue goals. Human factors must be at the heart of any and every autonomous vehicle programme. And before we get anywhere near autonomy, we need to get through the semi-autonomous phase. It’s a strange phase, said Chris Rockwell, CEO of Lextant: “Vehicles still rely on the driver to take over if things don’t work well.” Ironically, partial automation is often more stressful than fully manual driving, as drivers need to constantly monitor whether the vehicle is doing what it is supposed to. Wolfgang Bernhart of Roland Berger concurred: “OEMs must figure out how to keep the driver engaged so they can take over if needed but they can still use their time practically or for entertainment.” The ideal experience for most vehicle occupants using a Level 5 vehicle on a regular basis will be one that is unambiguous and predictable, noted Sam Abuelsamid of Navigant Research.
Self-driving car networks will need a vast supporting ecosystem
“We can use our network to handle all different types of mobility,” said Uber’s Lesh. “We haul people, food and freight. Essentially it’s the same underlying technology.” The food and freight might turn out to be the easier parts of the business. For Uber’s self-driving fleet – indeed, for any self-driving fleet – there will need to be a vast supporting network, or ecosystem, as Lesh called it. Sharing vehicles is great in theory, but it will be difficult to control people’s behaviour, and the support network could be very busy on the upkeep of potentially vandalised and soiled vehicles. Such a back-end ecosystem, also mentioned by Maven’s Jeff Shields, would support the fleet of self-driving cars with a fleet management system and a service and maintenance network, all geared towards providing what Lesh predicted will be “a great rider experience that will encourage people to adapt to this new technology.”
Parking will be confined to history
Even those of us who like driving hate parking, remarked Jon Demerly of Volvo-Autoliv joint venture Zenuity. “But what if your car could drop you off, and then go and park itself?” In the smart cities of the future, autonomous cars will take you to your destination and then, if not required elsewhere, head to the city limits to park. Inner-city parking will be a thing of the past, freeing up valuable real estate currently occupied by parking garages. The elimination of parking is often cited as a leading benefit of autonomous drive technology. Up to 30% of city traffic is made up of people simply driving around looking for parking, said Nathan Donnell of INRIX Parking Services, which results in 37,000 gallons of fuel wasted annually. It’s an incredible statistic, given that, as Donnell pointed out, “There are 250 million cars in the US, and a billion parking spaces.”
Autonomous cars need to be…unremarkable?
If an AV is nothing more than the next available robo-taxi, looks won’t matter – and looks are likely to be dictated by seat configuration and ease of ingress and egress, rather than sportiness and aerodynamics. As for the way it drives, unremarkable is the name of the game. When a good human driver is in control of the vehicle, there are generally few if any sudden movements or unexpected turns. AVs will need to accelerate, brake and steer smoothly and unremarkably. Whilst solutions are being sought for this issue, Navigant’s Abuelsamid suggested suspension calibration as just one of a host of ways of making the ride as smooth as possible. The smoother the ride, the better – you won’t be hailing a robo-taxi for its torque and acceleration, but for its ability to drive without inducing fear or motion sickness, noted Abuelsamid.
This article appeared in the Q2 2018 issue of Automotive Megatrends Magazine. Follow this link to download the full issue
