Waymo to Partner With Insurance Startup For Self-Driving Car Passengers

Jennifer van der Kleut

Waymo, Google’s self-driving spin-off company, is reportedly forming a partnership with an startup out of Danville, California to insure passengers of its future self-driving car-hailing service, according to multiple news outlets including the Wall Street Journal.

The startup, based in Danville, California, is called Trov, Inc. Trov CEO Scott Walchek said the company will be working with Waymo “behind the scenes” to insure riders for any medical expenses, property damage or lost property. Walchek also said that, in most cases, the passengers won’t even be aware, since passengers won’t be charged for the service.

Waymo previously said it hopes to launch a test of its driverless ride-hailing service — which will be similar to apps like Uber and Lyft but with a driverless fleet — with at least a few vans in the Phoenix Metro area by next year.

News Roundup: Alphabet Gets Approval for Its Dream ‘Digital District,’ Two States Push Forward With Driverless Car Testing, and More

Jennifer van der Kleut

Google’s Alphabet gets a green light to create a ‘digital district’ in Toronto, that includes driverless taxis

One of Alphabet’s spin-off companies, Sidewalk Labs LLC, has signed a major deal with Canada’s Waterfront Toronto to create a miniature “digital city” within the bustling metropolis, in the Quayside area of the Eastern waterfront. The district will take a stab at what the future of transportation looks like by featuring all manner of robotic mobility, including robot taxis, “driverless bike-like vehicles,” robotic delivery vehicles and even autonomous trash collection. Read more about Alphabet and Toronto’s plans from Bloomberg News.


GM, Cruise Automation to become the first to test self-driving cars in Manhattan

Officials announced this week that together, General Motors and their newly acquired partner Cruise Automation will be the first to test self-driving cars on public roads in the state of New York. The tests will begin in early 2018. Each test car — a Chevrolet Bolt — will have a pair of humans on board to ensure safety, and will employ Level 4 autonomous technology within a geofenced location. As the editorial staff of Ars Technica put it, GM and Cruise will have their work cut out for them, surely–“Manhattan’s roads are a hellish agglomeration of potholes, double- and even triple-parking, and pedestrian and vehicle traffic unlike anywhere else in the country. Gridlock is routine, and few quarters are given by other drivers before slamming on the horn in displeasure and disgust.” I guess we’ll have to wait and see. Read more from Ars Technica.


California may allow self-driving cars to be tested without humans in them by 2018

The Golden State is considering allowing self-driving cars to be tested on roads without humans inside them by the middle of next year, officials announced last week. Officials from the Department of Motor Vehicles proposed a new streamlined timeline for the regulations on Oct. 11, allowing a 15-day comment period from the public. The proposal is expected to be finalized by the end of this year, and approved by the beginning of next year. Then, human-less test cars could be hitting the roads by June 2018 or possibly even sooner, reports indicate. Read more from the Los Angeles Times. 

Image: A line of self-driving Chevy Bolt test cars / Credit: General Motors

News Roundup: Google Forms Self-Driving Car Company Called ‘Waymo,’ Uber Starts Testing Driverless Taxis in San Francisco Without Permission, and More

Jennifer van der Kleut

A look at some of the most interesting headlines to come out of the driverless and connected-car industries this past week:

Google spins out driverless car arm into its own company: Waymo

It’s finally happened–Google has officially launched its own self-driving car company, known as Waymo. Google officially announced the formation of Waymo this week on Dec. 13, 2016, along with the launch of Waymo’s own website. Using the same fleet the company has been testing in four U.S. cities over the past few years, Waymo says its next steps will be to start allowing people to test drive its self-driving cars “to do everyday things like run errands or commute to work.” Read more on Waymo’s website.


Uber starts testing self-driving taxis in San Francisco without DMV’s permission

The industry marveled when Uber rolled out self-driving cars in Pittsburgh earlier this year, but the government was less than happy when the ride-hailing giant began trying out autonomous taxis in San Francisco this week–without the DMV’s permission. Not to mention, some people are reporting seeing the driverless cars make traffic violations such as running red lights. News reports indicate the California DMV is currently trying to get Uber to halt the use of the driverless cars in California until testing permits are finalized. Read more from Business Insider.


Univ. of Iowa to use USDOT grant to research autonomous cars, pedestrians

The University of Iowa will share a $1.4-million grant from the U.S. Department of Transportation with four other institutions, and says one of the areas it will focus on with the research money is the safety and viability of self-driving transportation. Dr. Joseph Kearney, a computer science professor at the school, said “There’s work being done and work that will be done, that will look at interactions between pedestrians and bicyclists, and automated vehicles and semi-automated vehicles, in order to see how pedestrians respond.” Read more and see video footage from KCRG-TV.

News Roundup: A Semi-Autonomous Motorcycle, Driverless Cars Hit Public Roads in England, and More

Jennifer van der Kleut

A roundup of some of the most interesting news to come out of the driverless, connected-car world this week:

BMW says helmets won’t be needed with their self-balancing motorcycle

While most of the world is focused on semi-autonomous features that can make cars safer, BMW has been quietly focusing on a semi-autonomous motorcycle. This week, the auto manufacturer unveiled its design for the Vision Next 100 bike, with features like semi-autonomous steering and self-balancing wheels. Instead of a helmet, the bike will come with a visor that has an internal display super-imposed over the road and surrounding environment. The bike’s connected-vehicle system will give alerts about obstacles and risks on that display. BMW says the self-balancing wheels are so effective, a rider won’t even need to put their feet down on the ground when they stop, and it will be so hard to crash the bike, traditional helmets and padded, protective clothing won’t be necessary. Read more about the Vision Next 100 on CNNMoney.

Driverless cars tested on UK public roads for the first time

As Britain keeps moving toward its goal of having driverless cars on the road by 2020, a test car hit the public streets of Milton Keynes for the first time on Tuesday. Traveling at about 5 km per hour, the small two-seater driverless pod car navigated the streets of the largely pedestrianized southern town, stopping for people that crossed in front of it and safely turning corners. The pod car, heavily adapted from a compact Renault car, was developed by the Oxford University spin-out Oxbotica. Read more about the driverless car’s first public trip from Reuters.

Lots of driverless news out of California this week

According to news outlets like Ars Technica, Wall Street Journal and Elektrek, things are really heating up in California, where the number of companies that have been issued permits to test autonomous vehicles has just climbed to 17, up by three just since the end of summer. The two newest permits were issued to Wheego, an electric vehicle powertrain engineering company, and Valeo, a familiar name in the industry as a longtime tier-one automotive supplier. Also recently, Chinese tech firm Baidu received a testing permit. In other California news, Elektrek was one of the first to spot prototypes of Google’s long-awaited self-driving Chrysler Pacific mini-vans in Mountain View last weekend, and published a few somewhat grainy photos. Read more recent industry news from Ars Technica.

Image: Vision Next 100 semi-autonomous motorcycle prototype, by BMW.

News Roundup: Bentley Tests On-Demand Gas Fill-Up, Fly-Mode Wins With 3-D Printed Car and Drone Package, and More

A roundup of interesting headlines from around the driverless and connected-car world:

Bentley teams up to introduce on-demand gasoline fill-up service

Imagine your head is hitting the pillow one night, and you’re running through all the myriad errands you’re going to need to make in your car the next day. And then you remember, your gas tank is on empty. Well, if you lived in California and you owned a Bentley, that wouldn’t be a problem. It’s true, Bentley is teaming up with tech startup Filld to test out on-demand fuel delivery in California. With Bentley’s connected-car system, on-demand fuel is available 24/7, without even needing to hand over your keys. A technician will show up and refuel your car through the gas port, and a bill will be sent to your home. So, you could wake up the next morning to a full tank of gas, your car ready to hit the road. Pretty cool! Read more about this from AutoEvolution.

Check out this 3-D printed autonomous vehicle that features a scout drone

Imagine traveling with your car through gridlock traffic (not hard, I know). But then imagine you could send out your car’s drone to fly high overhead and scout out the best route to get around the traffic and get you to your destination faster. That could become a reality soon, thanks to Fly-Mode, a team of inventors that recently won the Essence of Autonomy Challenge, hosted by Local Motors and Mouser Electronics and judged by MythBusters star Grant Imahara.  The quad-motor drone has a landing pad on the rear of the vehicle, and occupants can use a joystick to control the drone, sending it up in the air to scout out surroundings, which Imahara says makes one feel as though they are in a flying vehicle, up in the air along with the drone. The drone’s images are sent to a projection screen inside the vehicle that can be inflated and deflated as needed. Read more about Fly-Mode from Popular Science.

Google inches closer to public release of self-driving cars, hits 2 million miles on public roads

Few companies have invested nearly as much time or money in the development of self-driving cars as Google’s Alphabet Inc. Recently, the company hit 2 million miles of driving on public roads. And in fact, it only took a little more than one year to go from 1 million to 2 million, as the fleet now consists of more than 60 self-driving test vehicles in four states. As news outlets report on the new benchmark, a reporter with Verge describes his recent ride in a Google self-driving car, including how he was a passenger while the car completed an obstacle course that had the vehicle dodge pedestrians, slam on the brakes as other cars cut in front, and navigate cyclists. Read more from Verge.

Silicon Valley Takes On Smart Mobility

Burney Simpson

Silicon Valley is in the middle of changing the way workers get to work. Then again, maybe not.

The Smart Mobility office of Joint Venture Silicon Valley (JVSV) seeks to implement several new transportation systems that will get people out of their cars, use public transit and bikes, and spew fewer greenhouse gas emissions.

This massive change relies in part on new technology developed by the growing autonomous driving industry.

The problem is that the Smart Mobility group doesn’t have the authority to force commuters to change. Further, the new technology is still being tested or has yet to find a large group of users to make it commercially viable.


Instead, Smart Mobility must rely on employers and staff to make the effort to get off their butts and make a difference. Employers have to see savings, or even revenue generation, while employees must be rewarded in some way.

No doubt the Bay Area has some serious traffic issues.

Congestion around San Francisco is ranked by Inrix as second worst in the nation, behind only Los Angeles and tied with Washington, D.C. San Francisco commuters wasted 75 hours sitting in traffic in 2015.

Steve Raney, executive director of the mobility group, believes that the best way to cut congestion and the resulting greenhouse gas emissions is to reduce Single Occupancy Vehicle (SOV) travel from 75 percent of travel in the Bay to 50 percent of travel.


Image by

That would mean cutting 1 million car trips a day from the current 3 million. If that happened it would eliminate 1.3 million tons of greenhouse gas emissions annually, according to a white paper from the mobility group.

Raney points to work done with Stanford University as one path to cutting SOV commutes.

Stanford began by charging its SOV commuters $3 a day to park at the school, a benefit it had previously offered for free. It then ‘feebated’ that revenue to its commuters that walked, biked, took a shuttle bus, used the Santa Clara Valley Transit Authority (VTA), or CalTrans, or the on-demand rideshare.

(A feebate is a system of charges and rebates designed to reward or penalize energy-efficient practices.)

As a result, Stanford reduced its SOV numbers from 75 percent of commutes to 49 percent, according to Raney. The school then saved the $107 million it had planned to spend on new parking facilities, he says.

In essence, Stanford replaced free parking with a shared travel benefit.

“Free parking for employees is a tradition,” says Raney. “(Employers) can phase in a charge of $3 for driving alone, or make it free to bike or walk. Just shift the dollar around for vehicles.”

The feebate concept was the most viable way to implement change when compared to such proposals as increasing the gas tax or imposing road user charges or implementing workplace parking charges, according to Raney’s research.

Feebates is just one aspect of the ‘Reducing Bay Area Commuting by 25%’ white paper from the JVSV mobility group.


Its Fair Value Commuting solution includes another four components, much of it relying on new technology associated with the developing autonomous driving industry:

  • The Enterprise Commute Trip Reduction (ECTR) is software from such firms as Luum and RideAmigos that allow employers to track staff commutes, and offer commute incents/de-cents.
  • Mobility Aggregation (MobAg) apps track and display multiple travel options – public/private transit, rideshare, carshare, bikeshare, van pool, etc. Vendors include Moovit, Transit App, Urban Engines, TripGo, Swiftly, Moovel, and Siemens. (In a perfect world, MobAg apps will also offer payment capabilities and update the ECTR so the employer can track employee activity).
  • Gap Filling is a catchall term for more transportation options and payment ideas, with a focus on first mile/last mile challenges. Gap fillers can range from low-income transit subsidies to Lyft/Uber peer-to-peer rideshare, escooters, public microtransit (VTA Flex3), private microtransit (Bridj, Chariot), private motorcoach (RidePal), telecommuting, and autonomous microtransit (EasyMile).
  • Systemic Obstacles refers to developing uniform payment systems and integrated routes by multiple transit agencies; developing an interoperable mobility software system.

The white paper concedes that much of this infrastructure is not in place. The MobAg apps are still gestating, the ECTR software needs to add features, and the public is catching up with all the gapfillers.

For example, payment apps do not allow for interoperability between the two dozen transit systems in the Bay area alone.

The Daimler purchase of smartphone e-ticketing firm Globe Sherpa might address this when its capabilities are combined with Daimler-owned RideScout. Still, this is a work in progress.


And even the feebate concept gets pushback despite its proven benefits. One Joint Venture collaborator notes that many employers don’t do feebates even when it may be in their best interest.

Raney is moving forward, saying that in September he will get an answer on four grant proposals that would allow him to expand his research and membership.

Meanwhile, he has drafted a proposal for the California legislature that would put a cap on free parking benefits an employer could provide to staff that commute in Single Occupancy Vehicles. The employer starts paying a fee when the percent of its SOV commuters tops a defined threshold.

America’s SOV mode of commuting once sold cars and built the wide-open freeways that took us to suburbia. Today it means congested highways filled with greenhouse gas delivery devices. It could take decades for the nation to break its SOV habit.


Photo by Coolcaeser.

Will Bolt EV Give a Jolt to GM Earnings Call?

Burney Simpson

The 2017 Chevrolet Bolt EV may include autonomous technology and Lyft drivers will get the first models off the production line, according to a flurry of news reports.

GM executives are considering addressing these topics during the auto OEMs second-quarter conference call on Thursday even though the Bolt EV won’t be officially released until late this year.

First, the electric-powered Bolt will include technology from Cruise Automation, the developer of self-driving systems that GM bought early this year, reports Motley Fool.

The Verge reported in May that Bolts outfitted with autonomous tech sensors were being live tested on San Francisco streets and Cruise co-founder Kyle Vogt was behind the wheel of one of them.

Second, the 2017 Bolt EV may be offered initially to Lyft drivers in California and Colorado through the ridesharing firm’s Express Drive program, Fortune disclosed. GM invested $500 million in Lyft in January.

Express Drive offers special rental prices on select GM vehicles to Lyft drivers who complete a certain number of rides per week. The Express Drive program has been over-subscribed since it was launched in Chicago and expanded to three more cities. Los Angeles and San Francisco may offer the program this fall.

Full production of the 2017 Bolt EV could begin in October according to Green Car Reports. Plans call for a base price of $37,500 and a production level of about 25,000 vehicles.


GM President Dan Ammann said at a Fortune conference this month that both rideshare and autonomous vehicles are interesting “and it gets really interesting if you put the two together.”

GM’s call to analysts to discuss second quarter results is scheduled to begin this Thursday at 10 a.m. Eastern time.

Along with revenues and earnings numbers, GM will tell analysts the price it paid for Cruise, a figure it has kept under wraps. Many press accounts estimated San Francisco-based Cruise cost GM $1 billion.

Proponents believe the 2017 Bolt’s 200-mile range will alleviate consumer concerns about running out of power. It is set to compete with two other electrics in the mid-$30,000 range – the Tesla Model 3 and the remade Nissan Leaf.

The Bolt may get the glory if it comes out first. The Model 3 has garnered about 400,000 reservations even though it won’t be put into production until the fall of 2017. The Leaf is scheduled for the 2018 model year.

Hype-less AVS 2016 Asks the Right Research Questions

Burney Simpson

Is this the show for autonomous driving nerds? No, that’s not quite fair. The Automated Vehicles Symposium 2016 is for the research types, the folks that dig deep to develop the systems that are going to bring driverless cars to fruition.

If that’s a nerd, wear the name proudly.

The AVS organizers must be doing something right. Last year it drew 870 attendees, a 50 percent rise from 2014. This year it looks to bring more than 1,000 academics, researchers and government staff to the San Francisco Hilton at Union Square on July 19-21, with ancillary meetings on the 18th and 22nd.

The conference is managed by the Association for Unmanned Vehicle Systems International (AUVSI), and the Transportation Research Board (TRB), a division of the National Research Council.

The Symposium calls itself the “largest gathering in the world of professionals involved with making automated vehicles a reality.”

Jim Misener acknowledges there’s a certain nerd factor to the show, but that’s because it is heavily weighted to engineers, scientists and deep thinkers.

“This show is less about the hype, and more about understanding what the research questions are,” said Misener, a director of technical standards with Qualcomm Technologies, the telecom giant’s chip design and R&D arm.


There are two parts to the AVS, notes Misener.

In the morning there are short speeches on big topics by the likes of U.S. Transportation Secretary Anthony Foxx, and reps from Ford and Nissan. Speakers cover ‘important matters’ like ethics, regulations, and autonomous activities around the globe.

After lunch you get the real action with 4-hour multipart breakout sessions that include presentations, Q&As, panels, videos, the ubiquitous PowerPoints, and general discussion. These sessions are designed to encourage frank interaction, and are closed to the media.

Misener helped to organize two of the 22 breakouts.

Enabling Technologies focuses on the foundational technology for driverless vehicles – mapping, algorithms, communications, sensing (sensors), and data.

In brief, the seminar is designed to educate on the strengths and limitations of each of these five technologies in 2016, and how they might work together in successful autonomous vehicle deployments. The discussion can lead to an understanding of technology gaps, and the research needed to close the gaps.

The ultimate goal is to define state of the art driverless technology, and determine how we can advance to that, said Misener.

He also helped organize a workshop on the aftermarket technology for autonomous vehicles.

AUVSI14aThe average car in the US is 11 years old, making aftermarket devices an important way to get autonomous technology into vehicles already on the roads.

“There could be a market for these devices. They could usher in safety and mobility services that get us to automated vehicles,” said Misener.

Other breakouts will address such topics as sustainability, cybersecurity, shared mobility, you get the drill, the usual.

Many attendees will come early for the Monday, July 18 ancillary meeting of the engineering organization SAE On-Road Automated Vehicle Standards Committee. The second ancillary meeting is on Friday, July 22, with the EU-US-Japan Automation in Road Transportation Working Group.

In addition, the AUVSI is holding on July 18 the Startup Connection at the Hilton. It offers demonstrations, presentations and networking for firms in unmanned systems and robotics, and investors looking for new companies.

Riders Get on the V2I Bus with Smart Stop App

Burney Simpson

A public-transit oriented smart phone application using V2I (Vehicle-To-Infrastructure) communications was tested successfully at the ITS-America 2016 conference in San Jose, setting the stage for further development.

The “Smart Stop” app uses Dedicated Short-Range Communications (DSRC) Wi-Fi technology to allow waiting bus passengers and buses to communicate with each other.

The Santa Clara Valley Transportation Authority (VTA) teamed up with Smart Stop developers Renesas Electronics America Inc. and eTrans Systems to test the app through three different demonstrations with 40-foot buses.

“The test went very well,” said Gary Miskell, chief information officer for the Authority. “This was the proof of concept. Now that we passed we can get funding (and move forward) on real development.”

The waiting passenger uses her smart phone or a kiosk touch screen to send a stop request with the Smart Stop app. That stop request informs the Santa Clara system she is at a specific stop waiting for a specific bus.

The stop request goes to a Road Side Unit that transmits it to the bus on-board unit which generates an audible and visual alert to the driver.

Smart Stop will notify the passenger through her smartphone or kiosk that the bus is approaching her stop.

For connected vehicle proponents the success of the test shows that DSRC can be used to make roads safer and more efficient. Smart Stop is an example of V2I technology that connects a fleet system with infrastructure (the Road Side Unit) by using Wi-Fi communications.

That’s great for the techies, but Miskell is looking to Smart Stop as something that can help him solve a day-to-day problem.

“Sometimes drivers don’t see the waiting person. But a stop request makes the driver stop,” he said.


That gives passengers a greater sense of control, which all transit riders appreciate, but it is “especially important for those with disabilities,” said Miskell.

That suggests that the app’s capabilities might be expanded to better serve passengers with special needs, such as those in wheel chairs and those with bikes.

But let’s not get ahead of ourselves, cautions Miskell.

In the near term he’d like to conduct a test of Smart Stop on select VTA routes with a limited number of passengers.

“We’ll get some customer feedback, track usage, and take it to the steering committee,” said Miskell.

If they like it, the VTA could put more funding behind the technology. And that’s when you get DSRC V2I technology solving day to day problems.

Driverless Vehicles: The Leading Nations for Testing Without a Human Driver

Anurag Maheshwary

Anurag Maheshwary is an attorney at a federal government agency. All viewpoints here are those of Mr. Maheshwary.

The race to develop the first driverless vehicles (i.e., NHTSA Level 4 and SAE International Standard J3016 Level 5) has spurred a fierce competition, not just among industry stakeholders, but also among jurisdictions vying furiously to establish themselves as the premier global test-bed for such technology. Although many jurisdictions have adopted a formal framework to permit testing driverless vehicles when a human driver is present, few, if any, have addressed testing without human driver presence. Over the past year, however, a frenzy of developments has occurred in jurisdictions worldwide that may soon enable driverless vehicles to be tested without human drivers. This survey highlights some recent developments in jurisdictions that have adopted a formal framework (comprising regulatory, legislative, and/or quasi-regulatory actions) that addresses public road testing or mass deployment of driverless vehicles without human presence. The survey compares and contrasts differing approaches among jurisdictions and ranks them according to their “testing friendliness” from the perspective of industry stakeholders. Finally, the survey concludes that although some U.S. states were the early global leaders, that lead is in danger of being eclipsed by other jurisdictions that are quickly catching up.

I. New Zealand: The World’s Most Testing-Friendly Jurisdiction — For Now

In what seems to be a well-kept secret, New Zealand — yes, New Zealand — appears to be the world’s most testing friendly jurisdiction. In February 2016, the New Zealand Ministry of Transport issued guidelines entitled, “Testing Autonomous Vehicles in New Zealand” that set forth prerequisites for stakeholders seeking to test driverless vehicles.[1] Among them are liability and indemnity insurance, submission of a “safety management plan” setting forth a description of the technologies being tested, a summary of prior private testing thereof, a log of incidents that occurred during testing, and compliance with (or application for exemption from) applicable regulations.[2] Upon satisfaction of these conditions, New Zealand will permit testing on any public roads without restriction. In contrast to other jurisdictions that have adopted traditional rulemaking or legislative tools to enable testing (see below), New Zealand has adopted a purely “quasi-regulatory” approach (e.g., government-issued guidelines, policy statements, and exemptions from vehicle standards).

nz2While New Zealand’s approach is unique, what truly distinguishes the country is its express proclamation that existing laws do not require a human driver to be present in the vehicle: “A particular advantage of testing autonomous vehicles in New Zealand is that our legislation does not explicitly require a vehicle to have a driver present . . . . So long as any testing is carried out safely, a truly driverless vehicle may be tested on public roads today.”[3] Granted, the guidelines are silent as to whether existing laws implicitly require human driver presence. However, the guidelines’ plain language and multiple references to testing without human driver presence show a clear intent to allow such testing.[4] Moreover, the guidelines encourage stakeholders to seek exemptions from provisions in vehicle standards “where requirements are clearly unreasonable or inappropriate,” suggesting a willingness to remove implicit barriers to testing driverless vehicles.[5] With respect to deployment, however, New Zealand has not proposed any enabling initiatives, opting instead to monitor international developments and respond when appropriate.[6]

New Zealand’s newly issued testing guidelines specifically encourage testing of driverless vehicles without human driver presence. Although there is an open question as to whether implicit regulatory barriers exist, New Zealand’s willingness to entertain exemptions suggests an openness to address that issue quickly. Arguably, for those seeking to test a truly driverless vehicle today, New Zealand is the most testing-friendly venue in which to do so.

II. U.K.: Runner-Up

After New Zealand, the U.K. appears to be the world’s second most testing-friendly venue. The U.K. has been aggressively developing a quasi-regulatory approach to facilitate testing of autonomous vehicles, publicly declaring its goal of becoming the global leader in this area. Since 2015, the U.K. has issued several policy guidance documents to achieve this goal. In February 2015, the U.K. Department for Transport issued a report entitled, “The Pathway to Driverless Cars: A Detailed Review of Regulations for Automated Vehicle Technologies.”[7] This report summarized findings of an audit of existing U.K. motor vehicle regulations, which was tasked with identifying regulations that were incompatible with the testing of autonomous vehicles. The audit report concluded that existing regulations permit such testing on public roads, provided that a human driver is present in the vehicle.[8] Importantly, no certifications, permits, or posting of bonds are required to commence testing, which are significant burdens imposed by many other jurisdictions. Finally, the U.K. committed to issuing detailed testing guidelines in collaboration with stakeholders. Notably, the U.K. touted its “light touch/non-regulatory approach” as “quicker to establish, more flexible and less onerous for those wishing to engage in testing than the regulatory approach being followed in other countries, notably in the U.S.”[9]

In July 2015, the U.K. issued its testing guidelines in a report entitled, “The Pathway to Driverless Cars: Code of Practice for Testing.”[10] Although non-statutory, the U.K. clearly intended the testing guidelines to be affirmative proscriptions, warning that non-compliance would constitute negligence.[11] The guidelines established detailed requirements for driver licensing and training, documentation of prior in-house testing on closed roads, data recording devices, data privacy and cybersecurity protections, among other things. But most critically, the guidelines appear to soften — or depart completely from — the audit report’s earlier requirement of human presence. The guidelines provide that either a human driver or a “test operator . . . who oversees testing . . . without necessarily being seated in the vehicle” shall supervise testing and be able to override autonomous mode at any time and take control of the vehicle.[12] This conflicts directly with the audit report, which provides that “[r]eal-world testing . . . is possible . . . providing a test driver is present and takes responsibility for the safe operation of the vehicle.”[13] It is unclear how to resolve this conflict. Still, given that the audit report predates and defers to the guidelines for establishing testing requirements, driverless vehicles without a human driver present arguably could be tested so long as a remotely located human supervisor is capable of taking control of the vehicle. As with New Zealand, stakeholders should consider seeking interpretations or exemptions to resolve this ambiguity.

In contrast to its quasi-regulatory approach to enable testing driverless vehicles without human driver presence, the U.K. recognizes that deployment would require regulatory and legislative actions, and the country has established a roadmap for doing so. By summer 2017, the U.K. aims to identify necessary regulatory reforms, and by the end of 2018, it would complete an assessment concerning the need for enabling legislation.[14] Overall, the U.K.’s testing guidelines raise the tantalizing prospect of testing driverless vehicles without human driver presence, pending resolution of conflicting guidance in the audit report. A resolution endorsing the view set forth in the guidelines could establish the U.K. as a leading testing-friendly jurisdiction in the near term, as well as a leading venue for deployment in the longer term.

III.      U.S.: Federal Leadership vs. State Patchwork Quilt 

At the state level, the U.S. was the pioneer in evangelizing autonomous vehicle testing. In 2011, Nevada became the world’s first jurisdiction to adopt a formal regulatory framework permitting testing on public roads. As of this writing, three states (Florida, California, and Michigan) and the District of Columbia have adopted formal frameworks explicitly permitting such testing, and many others are currently considering regulation and/or legislation. Collectively, roughly two dozen states have proposed more than fifty bills varying widely, resulting in inconsistency and confusion.[15] For example, Michigan restricts testing to OEMs, California requires the posting of a $5 million bond, and Nevada requires the presence of two licensed individuals in the vehicle, storage of pre-collision sensor data, special license plates, and geographic restrictions on testing.[16] This fractured approach has created a regulatory “patchwork quilt” that impedes testing, and it is largely due to an absence of comprehensive federal guidance. However, recent NHTSA announcements and actions aim to resolve this regulatory morass and create a uniform national policy. A summary follows.

A. Federal: A New, Uniform Approach May Soon Provide Leadership

In January 2016, Transportation Secretary Anthony Foxx unveiled the Department’s new strategy for autonomous (i.e., NHTSA Level ≤ 3 and SAE J3016 Level ≤ 4) and driverless vehicles, comprising various announcements and milestones that NHTSA committed to meeting this year. Among them, first, by July, NHTSA would propose guidelines regarding the deployment of driverless vehicles, along with a common understanding of requisite performance criteria and test procedures needed to assess them. Second, by July the agency would release a model state policy concerning both testing and deployment of autonomous and driverless vehicles that offers a path to consistent national policy. Third, NHTSA acknowledged that its existing legal authority is likely insufficient to support mass deployment of autonomous vehicles, “including those designed without a human driver in mind,” and stated that it would consider seeking additional authority via legislation or initiating rulemaking. Finally, Secretary Foxx encouraged stakeholders to seek FMVSS interpretations or exemptions, where appropriate, to facilitate testing or deployment of driverless vehicles — including those without human driver presence.[17] With these announcements, the U.S. has committed to a comprehensive framework to accelerate the advent of truly driverless vehicles.

Next, in a February 2016 letter, NHTSA responded to a rule interpretation request from Google concerning the definition of the term “driver” for purposes of compliance with certain FMVSS provisions.[18] In a major policy shift that may have a significant impact on enabling driverless vehicles, NHTSA interpreted “driver” broadly to include an artificial-intelligence self-driving system, which opened the door to legal recognition of truly driverless vehicles for the first time.[19] However, the agency cautioned that until it develops performance criteria and test procedures for evaluating whether a self-driving system meets various FMVSS provisions, it could not categorically interpret self-driving systems as compliant with existing regulations.[20] The agency allowed that some FMVSS provisions could be addressed in the short term via requests for interpretations and petitions for exemptions, but cautioned that other provisions, such as requirements for manual controls, would require a traditional regulatory solution.[21]

In March 2016, the DOT issued a preliminary report entitled, “Review of Federal Motor Vehicle Safety Standards (FMVSS) for Automated Vehicles,” which summarized results of an audit of existing FMVSS provisions.[22] One purpose of the audit was to identify standards that include references, both implicit and explicit, to a human driver that could impede deployment of driverless vehicles.[23] The audit report painstakingly examined each of the 73 standards specified in the FMVSS and identified specific provisions that could be problematic. Among other things, the audit found that although few existing regulations would hinder deployment of autonomous vehicles designed for human driver presence, vehicles without human drivers or manual controls would indeed conflict with FMVSS provisions or policy objectives.[24] However, the DOT noted that its analysis predated NHTSA’s interpretation of the definition of “driver” and therefore did not reflect this development.[25]

Together, imminent federal guidance documents, results of the regulatory audit, and NHTSA’s innovative interpretation of “driver” show leadership that may establish the U.S. as among the world’s preeminent venues for both testing and deployment of truly driverless vehicles. In the near term, NHTSA’s quasi-regulatory approach should help harmonize conflicting state approaches, resulting in a consistent and uniform national policy. Longer term, contemplated legislative and rulemaking efforts should facilitate deployment.

B. States

  1.    Florida: Soon to Become the Leader?

In 2012, Florida became the second state (after Nevada) to permit operation of autonomous vehicles on public roads. However, it placed onerous restrictions on those wishing to do so, including: restricting operation to testing only, restricting testing to either OEMs or educational institutions, requiring the presence of a human driver, and requiring either proof of insurance or a bond in the amount of $5 million.[26]

Florida_topographic_map-en1In April 2016, however, Florida enacted H.B. 2027, which was a sweeping expansion of the permissible uses of autonomous vehicles on public roads.[27] Among other things, the legislation jettisoned all of the foregoing limitations — including the requirement of human presence. Theoretically, removal of this limitation would enable driverless vehicles without human driver presence to be tested or deployed on Florida roads. It remains to be seen, however, whether implicit Florida or federal regulations would thwart such testing or deployment. Still, if stakeholders are able to obtain favorable regulatory interpretations or exemptions, Florida may have catapulted itself to the status of being among the world’s most testing-friendly venues. The legislation takes effect on July 1.

  1. California: Soon to Become the Laggard?

California is one of the world’s only jurisdictions to develop regulations for the deployment of autonomous vehicles. In December 2015, California released draft rules that address only the transition from testing to deployment, but not full deployment.[28] The rules would require, among other things, manual controls, human driver presence, new safety and performance requirements, certifications from third-party testing organizations, regular submission of usage reports, and restriction of distribution to leases, not sales.[29] Unsurprisingly, many stakeholders have criticized the proposed rules harshly as unnecessarily restrictive and likely to stifle development of autonomous technologies. In particular, critics have condemned the requirements for human drivers and manual controls, contending that if finalized, such restrictions would deter firms from deploying — or even testing — driverless vehicles in California.

Echoing these concerns, in February 2016, legislators introduced bill A.B. 2866 in the California State Assembly, which amounts to a full frontal assault on many of these restrictions in the draft regulations.[30] Among other things, the bill would overturn the proposed requirements for human presence and manual controls, and it directs California, by July 2018, to adopt regulations expressly permitting the testing and deployment of driverless vehicles without human drivers or manual controls. The bill would also expressly permit public road testing and deployment of driverless vehicles upon adoption of the new regulations.[31] The bill is currently in the embryonic stages of the legislative process. Until it is enacted, however, California’s strict regulatory approach has brought with it the dubious distinction of being the least hospitable jurisdiction in this survey in which to test truly driverless vehicles.

IV. Australia: The Most Comprehensive Approach Yet

In November 2015, Australia launched a comprehensive plan to enable testing and deployment of autonomous and driverless vehicles. The plan sets forth specific milestones, including auditing all relevant federal and state laws and regulations to identify potential barriers to testing and deployment, recommending remedial actions, and commencing some preliminary reforms. Australia has set an aggressive timeline to achieve these milestones — just over one year.[32]

Reliefmap_of_Australia1The first fruit of this effort was borne in February 2016, when Australia’s National Transport Commission (“NTC”) issued an initial report entitled, “Regulatory Barriers to More Automated Road and Rail Vehicles Issues Paper.”[33] The report provided an overview of Australia’s existing regulatory landscape and identified key issues that could impede testing and deployment of autonomous and driverless vehicles. The report concluded that although most aspects of its existing regulatory framework are unlikely to impede driverless vehicles, still other potential barriers exist. In particular, the NTC noted that “the most significant barrier[s]” were some federal and state laws, as well as vehicle safety and performance standards, that implicitly require human driver presence.[34] The report identified several additional categories of potential barriers, including: ambiguity in the definition of “control” concerning who or what is in “control” of an autonomous vehicle, liability issues, and data privacy and access thereto.[35] The NTC painstakingly examined the proper role of government in removing the foregoing barriers, and considered whether the optimum approach should involve rulemaking, legislation, or a quasi-regulatory approach. Ultimately, the NTC cautioned against a rush to rulemaking or legislation at this time, fearing that such processes could chill innovation in this rapidly evolving technological area, devolving into a U.S.-style regulatory patchwork quilt.[36]

In May 2016, the NTC issued a follow-up report entitled, “Regulatory Options for Automated Vehicles.”[37] Analogous to audits conducted by the U.K. and U.S., Australia audited relevant laws and regulations and identified specific provisions — 716 in total — that were potential barriers to testing and deployment.[38] It also proposed recommendations to address these issues, which are subject to revision based on upcoming public comments. Importantly, these reforms would be coordinated at the federal and state levels, in contrast to the current approach in the U.S. The remedies differ in purpose and scope and would phase-in over five years. Phase 1, which would commence as soon as possible, would facilitate testing of driverless vehicles and deployment of autonomous vehicles by introducing national guidelines to support a uniform approach to testing and state-based exemptions to existing standards, as well as clarifying the meaning of “control.” Phase 2, which would commence within two years, would enable deployment of driverless vehicles via new legislation expanding the definition of “driver” to include autonomous driving systems, providing a “safety assurance framework” for autonomous vehicles, among other things. Finally, Phase 3, which would commence within three to five years, would initiate rulemaking to remove implicit references in vehicle standards to human drivers and manual controls, and issue new standards governing cybersecurity, data privacy, and other issues.[39] The NTC will make final recommendations to federal and state transportation ministers in November 2016, and in December 2016, the agency will issue a final report summarizing the direction chosen by the ministers.[40]

Australia has embarked on an ambitious journey that may establish it as a preeminent jurisdiction in which to both test and deploy truly driverless vehicles. Under its proposed regulatory reforms, public road testing of vehicles without human drivers could commence almost immediately, and full deployment may be possible in 2020. Like the U.S., Australia has embraced the idea that the term “driver” in vehicle standards should not be narrowly confined to humans, and instead should be broadened to include autonomous systems — a critical recognition made by no other major jurisdictions. However, unlike efforts in the U.S., Australia’s blueprint proposes comprehensive, coordinated reforms at the state and federal levels, which should provide greater uniformity and clarity. Coupled with Australia’s phased approach to reforms — ranging from a quasi-regulatory approach in the short term, to legislative amendments in the medium term, to initiating rulemaking in the long term — Australia’s overall framework is an innovative solution that should provide clarity and predictability to industry stakeholders seeking to test or deploy truly driverless vehicles.

V. Canada: Last Among the G7

Among G7 nations, Canada is farthest behind in developing a formal regulatory framework for testing autonomous vehicles and has no plans specific to driverless vehicles.[41]  (Canada is included in this survey due solely to its geographic and economic proximity to the U.S.)  However, recent developments show that Canada is keen to shed that status.  In January 2016, Ontario established a pilot testing program, making it the first province to enable public road testing of autonomous vehicles.[42]  Under the 10-year pilot, a human driver must be present in the vehicle, and a certification of insurance of $5 million is required; no special permits, license plates, identifiers, or other onerous conditions are imposed.[43]  Notably, the application process is highly streamlined, consisting of an online, single-page form that does not require submission of any supporting documentation.[44]  Applicants are simply required to certify, by clicking a checkbox on the form, that they have reviewed the requirements set out in applicable regulations.[45]  This simplified online form is a unique and innovative feature that should facilitate participation in testing.  Although Canada has not yet adopted a formal framework specific to driverless vehicles, Ontario’s actions have already catapulted itself ahead of many U.S. states with respect to testing autonomous vehicles and may hasten development of a federal approach for testing truly driverless vehicles.

VI. Conclusion

An analysis of recent international developments to facilitate testing driverless vehicles without human driver presence shows a continuum of widely varying regulatory approaches. At opposite ends of the spectrum are New Zealand, with its freewheeling, quasi-regulatory approach, and California, with its rigid, traditional rulemaking approach. Occupying the white space in the middle are the U.K., U.S. (federal), and Australia, which are following hybrid frameworks, consisting of a quasi-regulatory approach in the short term, with legislation and rulemaking contemplated in the longer term.

For those seeking to test a truly driverless vehicle today, New Zealand (and perhaps Florida beginning in July) is the most testing-friendly venue in which to do so, although there remains an open question concerning the existence of potential implicit regulatory barriers. The U.K. appears to be the runner-up, as it ostensibly permits testing by a remotely located driver, but a conflict in the relevant guidance documents needs resolution. In the U.S., a regulatory patchwork quilt at the state level, together with recent federal pronouncements, create much uncertainty. However, if imminent federal guidance resolves the regulatory quagmire, the U.S. may be poised to lead. Finally, the Land Down Under is perhaps the most interesting. Although Australia’s legal landscape bans testing vehicles without human driver presence, its proposed reforms, which comprise a comprehensive, coordinated effort at the federal and state levels, are broader and deeper in scope than those of the U.S. and ultimately may yield the most promising outcome for stakeholders seeking to test truly driverless vehicles. Overall, although much uncertainty persists from developments over the previous twelve months, the next twelve are certain to be even more momentous.

Anurag Maheshwary is an attorney at a federal government agency.  Previously, he was an associate at White & Case LLP, and a law clerk in the U.S. District Court for the Middle District of Florida.  Prior to his legal career, Anurag was an automotive engineer at Freightliner Corp.  Anurag holds a J.D. from Boston University School of Law, an M.S. in Mechanical Engineering from Virginia Tech, and a B.S. in Mechanical Engineering from North Carolina State University.  He can be contacted at

This article first appeared in the American Bar Association’s Transportation Quarterly.

Photo – Supreme Court by Brittany Hogan, 2005.

Relief map of Australia by Hans Braxmeier, 2008.


[1] N.Z. Ministry of Transp., Testing Autonomous Vehicles in New Zealand (Feb. 2016),

[2] Id. at 3-5.

[3] Id. at 1.

[4] See id. at 1 & 5.

[5] Id. at 4.

[6] N.Z. Ministry of Transp., Intelligent Transport Systems Technology Action Plan 2014-18, at 25 (May 2014),

[7] U.K. Dep’t for Transport, The Pathway to Driverless Cars:  A Detailed Review of Regulations for Automated Vehicle Technologies (Feb. 2015), [hereinafter U.K. Audit Report].

[8] Id. at 10.

[9] Id. at 11.

[10] U.K. Dep’t for Transp., The Pathway to Driverless Cars:  A Code of Practice for Testing (July 2015), [hereinafter U.K. Testing Guidelines].

[11] U.K. Audit Report, supra note 7, at 118.

[12] U.K. Testing Guidelines, supra note 10, at ¶ 2.11.

[13] U.K. Audit Report, supra note 7, at 10.

[14] Id. at Annex D, Rows 17 & 21.

[15] See Autonomous/Self-Driving Vehicles Legislation, (last visited May 26, 2016).

[16] See id.

[17] Press Release, U.S. Dep’t of Transp., Secretary Foxx Unveils President Obama’s FY17 Budget Proposal of Nearly $4 Billion for Automated Vehicles and Announces DOT Initiatives to Accelerate Vehicle Safety Innovations, DOT Press Release 02-16 (Jan. 14, 2016),

[18] Letter from Paul A. Hemmersbaugh, Chief Counsel, NHTSA, to Chris Urmson, Director, Self-Driving Car Project, Google (Feb. 4, 2016),–%20compiled%20response%20to%2012%20Nov%20%2015%20interp%20request%20–%204%20Feb%2016%20final.htm.

[19] Id.

[20] Id.

[21] Id.

[22] Dep’t of Transp., Review of Federal Motor Vehicle Safety Standards (FMVSS) for Automated Vehicles (Mar. 2016),

[23] Id. at 20.

[24] Id. at ix.

[25] Id.

[26] Fla. Stat. § 316.86 (2015).

[27] H.R. 7027 (Fla. 2016) (as passed by House, Mar. 11, 2016).

[28] Cal. DMV, Autonomous Vehicles Express Terms (proposed Dec. 16, 2015),

[29] See id.

[30] Assemb. B. 2866 (Cal. 2016) (as referred to Comm. on Appropriations, Apr. 19, 2016),

[31] See id.

[32] Austl. Nat’l Transp. Comm’n, Regulatory Barriers to More Automated Road and Rail Vehicles Issues Paper 5 (Feb. 2016), [hereinafter Austl. Preliminary Report]; Austl. Nat’l Transp. Comm’n, Regulatory Options for Automated Vehicles 9 (May 2016), [hereinafter Austl. Audit Report].

[33] See Austl. Preliminary Report, supra note 32.  The NTC is an independent, inter-governmental agency that develops and submits recommendations to the Transport and Infrastructure Council, which comprises transportation ministers from the federal and state levels.

[34] Id. at 7.

[35] Id. at 7-10.

[36] Id. at 22-24.

[37] Austl. Audit Report, supra note 32.

[38] Id. at Annex,

[39] Id. at 9-12.

[40] Id. at 25.

[41] Canadian Automated Vehicles Ctr. of Excellence, Preparing for Autonomous Vehicles in Canada 5 (Dec. 16, 2015), http://

[42] News Release, Can. Ministry of Transp., Ontario First to Test Automated Vehicles on Roads in Canada (Oct. 13, 2015),

[43] Ont. Regulation 306/15 (2016),

[44] Gov’t of Ont., Automated Vehicle (AV) Pilot Application Form,$File/023-5084E.pdf.

[45] Ont. Regulation 306/15 (2016),


Nothing Found

Sorry, no posts matched your criteria