2017’s prospects of autonomous, mobility and connectivity, explored by Susan Kuchinskas. [Mob.Kuchinskas.2016.10.12]
Continued from Part I
3. Legislation and regulation
Industry players have got their regulation story straight: guidance, not rules. As was seen in the kerfuffle about the California Department of Motor Vehicle (DMV) draft regulations, the industry is willing to put plenty of muscle into pushing back against attempts to regulate it.
In December 2015, the California DMV released draft regulations for autonomous cars that required a licensed human to be at the wheel, ready to take over in an emergency.
That sounded reasonable in theory, according to Elliot Katz, global company-chair of the connected and self-driving car practice of the law firm of DLA Piper but, meanwhile, the so-called machine-to-human handoff was looking more and more problematic. NHTSA and Virginia Tech’s tests showed that drivers of cars with Level 3 automation took an average of 17 seconds to accomplish the handoff.
In September 2016, NHTSA released its Federal Automated Vehicles Policy with a list of 15 safety assessments that manufacturers can voluntarily submit. “It clearly said that a fully autonomous vehicle should be driven entirely by itself and require no licensed human driver -- to the applause of industry. That enables Google and others to continue their development of self-driving cars with no steering wheels and no brake pedals.
“People refer to these as self-driving cars,” Katz says, “but really, these companies are designing a driver.”
In early October, the DMV backed down from its stance, allowing cars without steering wheels or human drivers to be tested on public roads, conforming to NHTSA’s guidance. However, it went on to make NHTSA’s voluntary 15-point checklist mandatory – and required carmakers to get permission from local authorities before testing. The Self-Driving Coalition for Safer Streets, founded by Google, Ford, Uber, Lyft and Volvo and fronted by ex-NHTSA honcho David Strickland, blasted the DMV, claiming that these regulations would slow innovation. (The coalition is also fighting with the City of Chicago.)
“The ethos in the Bay Area is, innovate before you regulate. On the flip side, there has to be some regulation in place. If it’s the Wild West, you won’t get buy in from the public,” Katz says.
NHTSA has shown itself to be generally on-board with this theory. Sam Abuelsamid, a senior analyst with Navigant Research, says that after publishing the guidelines, NHTSA seems content to be mostly hands-off. “They realise this technology is moving so quickly that it’s probably premature to do any kind of motor-vehicle safety standards. They are not in any rush to move ahead with full regulations but, if we do start to see more crashes, that situation could change.”
California seems to be trying to skirt the line between keeping the Silicon Valley economic engine revving and addressing the concerns of groups like Consumer Watchdog that are pressing to keep self-driving vehicles off the roads for now. “The reason that the industry reacted so harshly is that, in an ideal world, you’d have the Feds step in and say, ‘This is law of the land,’” Katz says. California’s proposal creates three different layers of regulations: federal, state and local. “It’s a lot for an OEM or new entrant to deal with.”
Gail Gottehrer, a partner in Axinn, a law firm that specializes in antitrust and intellectual property litigation, isn’t sure how helpful NHTSA’s guidance really is. Some in the industry expected it guidance to offer specifics about things like testing autonomous vehicles; instead it offered general principles and ideas, she says. “My sense is that NHTSA built it up into more than it turned out to be.”
On the other hand, she notes that the agency guidance very clearly demarcated its own role vis a vis that of the states. “NHTSA is asserting its muscle and telling states how broadly it sees its own role being in this autonomous space,” Gottehrer says.
Nevertheless, she says the states have much to figure out on their own: how will autonomous vehicles be licensed and how will they be marked in a way that lets other drivers or the police know their status? Also, will the states need to come together with some kind of universal licensing system, perhaps a different coloured license plate?
Gottehrer says: “In fairness to NHTSA, that’s why they didn’t go further. There are so many branches of complexity for each aspect. There is wisdom in the iterative approach, to put out gen guidance, get feedback from players in the industry, and then move forward from there. It’s a huge challenge.”
Anne Teigen, a programme principal/attorney who focuses on transportation issues for the National Conference of State Legislatures(NCSL), an organisation that provides research and other services to state legislators and their staff, notes that 16 states passed some sort of legislation relating to autonomous vehicles in 2016. Most of them concerned testing; Utah simply directed state agencies to look into the whole thing and report back.
“The issue is that you have technology that is moving so fast and the technology hasn’t existed before. It’s hard to think of a regulatory scheme when we don’t know what technology is really out there,” Teigen says. To date, 34 states and Washington, DC, have considered such legislation, according to the NCSL.
Autonomous vehicles may ultimately reduce state revenue, according to Brookings’ Local Government 2035 Report, owing to the reduction in traffic tickets, towing fees and, if ride hailing and car sharing continue to grow, vehicle registration fees. However, in the short term, friendly state governments could see booming business.
That’s Michigan’s plan. Governor Rick Snyder is set to sign wide-ranging legislation that allows for testing on all public roads; supports mobility programs; and exempts vehicle-repair centres from product liability when they repair autonomous vehicles. Silicon Valley may have Google but Novi will have Google’s 53,000-square-foot self-driving car development centre – and that’s likely just the start of a new automotive research boom for the state.
4. Connectivity of the future: 5G v DSRC
What will be the connectivity platform for the future? Will cars equipped with DSRC constantly exchange information with each other and with smart infrastructure to avoid collisions and speed traffic? Or will 5G communications handle those tasks while letting autonomous cars access real-time maps and information while their passengers watch movies?
Opinion is split in the auto industry on whether we need DSRC as part of semi-autonomous or autonomous driving, or whether 5G will supplant it. The GSMA is touting 5G as the path to a new era in connectivity, with mobile networks that are “amazingly responsive, extraordinarily dense, blazingly fast … and very low power.” That’s the basic premise of the 5G camp: why settle for the limits of DSRC when 5G could do so much more?
On the other hand, DSRC is a full-fledged technology, while 5G isn’t even a spec, as Roger Melen, senior advisor, Toyota InfoTechnology Center USA, points out. For safety-related functions, he says, DSCR “is something that’s proven and earned, not just claimed.” In order to use a particular technology in cars, he adds, it must be tested in all kinds of circumstances, which DSRC has been.
Melen argues, 5G “doesn’t have a specification, so we really don’t know what it will do. It’s like that mythical beast that is going to solve all of your problems”. He notes that, in addition to the lack of any specifications, 5G would have shorter range than DSCR; moreover, it’s predominantly a line-of-sight system, while DSRC could operate beyond the range of a car’s sensors and can detect objects out of the line of sight, such as an obstacle hidden by another vehicle. Melen thinks the United States will mandate, at the least, that DSRC beacons be installed in cars by 2020.
Toyota has, of course, is betting on DSRC and rolling it out on vehicles in Japan. In 2015, General Motors (GM) became the first US automaker to commit to using DSCR for vehicle-to-vehicle (V2V) communications and it’s rolling it out in the 2017 Cadillac CTS. Harry Lightsey, GM executive director of public policy for emerging technologies, said that NHTSA’s proposed rulemaking on V2V technology is an example of a development that would help the company accelerate its strategy. “GM believes V2V technology holds tremendous potential to enhance vehicle safety,” he says. It’s true that 5G is rather far in the future; on the other hand, many posit that it will likely take until the 2025 or so for DSRC to penetrate enough into the installed base of cars that it would provide benefits – at which time we might see 5G.
At a recent conference, Gaurav Bansal, a senior researcher in cooperative automated driving at Toyota InfoTechnology Center, said that Toyota expects 5G using the millimetre wave frequency band to be in deployment between 2020 and 2025. By 2025, a car equipped with 5G would have a decent chance of meeting another such vehicle and exchanging information to be used for merging and non-line-of-sight views. “By 2030,” he said, “we will have sufficient deployment and the true potential of cooperative automated driving will be available.”
5. The view for 2017
Two interrelated things loom on the horizon: the possibility of more accidents involving autonomous vehicles and the certainty of more legislation regarding them. How that legislation shapes up may depend on how well self-driving cars can prove themselves on public roads.
Karl Brauer, senior director of content and executive publisher, Cox Automotive, is worried about carmakers moving too fast and releasing vehicles with full autonomous modes when they really shouldn’t. Pointing to Tesla’s Florida fatality, he says: “I don’t think we’ve seen the last of that. People will make assumptions about the capabilities of these cars and not pay attention. We will see more and more accident claims where the driver says, ‘It was the car’s autonomous system, not me.’ And will be hard for manufacturers to say, “No, it was you.’ As long as you having a discussion about who is in control at any given moment, you are opening things up for the wrong entity to have been in control.”
2017 will be a big year for legislation, according to Anne Teigen, a program principal/attorney who focuses on transportation issues for the NCSL, an organisation that provides research and other services to state legislators and their staff. One impetus is NHTSA’s release of its guidelines which spelled out the responsibilities of states versus the Feds.
Teigen says: “Right now, a lot of states are looking at defining what an autonomous motor vehicle means – things that are really basic.” She adds that most states understand that the industry doesn’t want to be too heavily regulated, because companies don’t know what technologies they will want to test. They’ve argued in the press to keep these regulations broad. “Legislatures get it,” she says.
However, they are getting pressure from some consumer groups and, possibly, some localities that are wary of the tech and want to slow things down. Volvo has moved to ease liability fears by promising to take full responsibility for accidents involving it self-driving vehicles, and Michigan’s new legislation specifically exempts repair shops from liability if they repair to manufacturers’ specs.
Connected home and IoT
Smart-home technology has supposedly been on the cusp of widespread penetration for more than a decade. Will this be the year it takes off? Ted Cardenas, director of business development and corporate planning at Pioneer Electronics, says: “It’s reasonable to think that in the next few years, there will be a real capability of being able to drive within a certain geofenced proximity of your home and have the security system and lights go on, so that when the garage door opens, the alarm doesn’t go off.”
He believes there will be more integration between a variety of connected devices in the home and the connected car, including not only home systems but also entertainment and information. This could be driven by Google and Amazon, both of which have a big stake in keeping consumers connected to content and information and their latest devices, Amazon’s Alexa and Echo, and Google Home, could be the spark.
Sudipto Aich, a principle researcher at Ford Research & Innovation Center, sees huge opportunity in the growing ubiquity of connection. He’s starting to see the immense activity in the Internet of Things space dispersing into the connected home – and Ford’s partnership with Amazon is an example. If Amazon devices can become part of production vehicles, he says, “Seeing customer interactions will lead to new paradigms of how people spend their time at home and in vehicles.” While the space is still very fragmented, he says, Ford’s work with connected-home companies will hopefully allow it to figure out how a fully connected world will fit into someone’s lifestyle.
Yet, car companies are still wrestling with what to do with the connections in their cars, Strategy Analytics’ Roger Lanctotsays. Whether they will be able to sell car data to others is unclear. But he foresees more creative uses of the on-board modem.
“Car companies will get more creative about what they’re doing with the connectivity in the car to transform the driving experience,” he says. Examples are delivering Waze-like functions directly to the car so that drivers can share information with others.
He thinks the industry will get a little closer to understanding how to use data to keep the connection with the customer intact after purchase – just a little. “We’re talking incremental,” he says.
Expect plenty of cash to be thrown around in 2017, as automakers and tech giants jostle to carve out positions in ride hailing, car sharing and other mobility-on-demand services including on-demand fleet services. But the true impacts of mobility schemes on the environment, public transportation, car ownership and carmaker profits won’t be clear for many years to come.
Traditional automakers will tout their ADAS and autonomous endeavours with varying degrees of caution or aggression. The dates for producing autonomous vehicles may be pushed back to 2022 or beyond, and claims will become more specific, in line with what Ford has said.
States that welcome autonomous vehicles to public roads will get an economic advantage, as well as public relations benefits. States that want to be more cautious than the industry wants will face backlash in the press and lobbying from industry groups like the well-endowed and rather disingenuously named Self-Driving Coalition for Safer Streets. And, like the State of California, they may have to back down.
Expect more drama between ride-hailing services, cities and drivers. The UC Berkeley Transportation Sustainability Research Center and the Natural Resources Defense Council plan to release a study in the spring examining the impact of ride-hailing on traffic in San Francisco. If it shows that they make traffic worse, cities may consider limiting their services or banning them outright.
Engineers with expertise in AI, computer vision, sensor fusion and the other technologies needed for autonomy will continue to be treated like rock stars, and they’ll have plenty of opportunities. Executives will move fluidly among automakers, Tier 1s, consumer electronics, academia and digital media. Promising start-ups will be acquired at eye-popping valuations. It will be a good year.
19 Apr 2017 - 20 Apr 2017, LONDON, UK
This is the largest and most comprehensive forum in Europe dedicated to connected car insurance. With over 400 attendees from across 25+ countries globally and 50+ executives speakers, this event will address the technological challenges and new business models to impact the motor insurance industry as it converges with the connected car.