Governments Must Help Consortia Drive Autonomous Progress

There are a number of consortia being created to push the way for us all to see connected and autonomous vehicles (CAVs) on our roads.

Nissan, for example, has joined ServCity to develop a blueprint to enable cities to introduce them into cities across the UK. Similar consortia have been created in the country, and internationally – such as the Autonomous Vehicle Computing Consortium (AVCC) in the United States.  The AVCC is consortium includes companies such as Arm, Bosch, Continental, Denso, General Motors, Nvidia, NXP Semiconductors and Toyota. The consortium’s website says their goal is to “accelerate the delivery of safer and affordable autonomous vehicles at scale” with a view to making fully autonomous vehicles a reality through “industry-level collaboration”.

In order to achieve this ambition the AVCC, like many other consortia including ServCity, have got off the starting block by developing a set of recommendations and defined standards to follow, with the AVCC focusing on a “system architecture and a computing platform to promote scalable deployment of automated and autonomous vehicles”.

Consortia trend

Paul Campion, CEO, of the UK’s Transport and Research Laboratory (TRL) explains why there an increasingly prevalent trend to create consortia for the development and testing of connected and autonomous vehicles: “Prof. Mariana Mazzucato, director of the Institute of Innovation in Public Purpose at UCL, has become quite influential in the UK government’s thinking. She talks about how innovation happens in the real world, and challenges the view that innovation occurs through investment capitalists alone.

“For example, in the iPhone the LCD display and the notion of the internet as well as a lot of the original research on semi-conductors came out of public expense. The idea that’s become common is ‘mission-based’ innovation, such as the Apollo Moon program. You need academia, industry and the state. The state is the only player that can afford a long-term view. Industry has to focus on a short-term view to make profits.”

He also mentions venture capitalist William Janeway, a partner in Warburg Pincus, who believes the state has to create the conditions for capitalism to be entrepreneurial. Essentially his thinking suggests that by encouraging the different public and private sector parties working together, transformation and innovations can be achieved. Campion says this principle is now well established, and so it applauds its use as a piece of UK government policy as “a really encouraging and positive statement about the UK’s industrial strategy”.

“The UK has become one of the best places to test CAVs. There has been a conscious and deliberate effort by government to work with industry to foster the ecosystem. An autonomous car is a dumb car with a yet to be developed supercomputer within it. Automation in vehicles requires a range of skills from production engineering to systems integration, as well as an understanding of the motoring environment, and there is a need to influence new regulations as well as the development of new AI and ML techniques. There isn’t one company that has all of these skills.”

Sophisticated CAV ecosystem

With consortia, a sophisticated CAV ecosystem can be created as has happened in the UK. These ecosystems involve inter-dependent companies. Some will be assembling components from specialized firms and there will be a wide range of other experts because he rightly comments that no single company could afford to invest in them.

Campion adds that there is also a need to understand the interest of insurers, which can only be achieved by sharing data and by comprehending the technical angle, the societal view in the context of a public-private relationship to drive innovation. To achieve their ambitions consortia need to develop blueprints for achieving the realization of us seeing CAVs going from the testbed to being purchased and used as a common form of mobility on public roads. Without a blueprint it could be hard to establish how to deploy these vehicles safely and correctly on city roads, while also creating standards for best practices that can be shared with other projects around the world.

Demanding commitment

Vishal Singh, CTO of Zenduit comments that there is a need for blueprints to enable cities to introduce CAVs to “demand attention and commitments from the various players”. He adds: “From each player, it must be clear that these key points are addressed: information on the vision of the project including mission, the challenges of CAV or risks that are present, the opportunities, the needs required to progress, the strategies to move forward, the cost and funding requirements, partnership opportunities, educational requirements, immediate actions required, time frames, performance measures, and key points of contact on the project.”

Like Campion, he believes it is critical for governments to support consortia: “Without their support, key government-based initiatives may not be successful and key players will look elsewhere for development opportunities. What this refers to is the requirements, regulations and infrastructure set out by governments must allow for consortia to feel confident in engaging with them to progress.”

In effect, he says, consortia are required for governments and cities to progress with Vision Zero initiatives, which are about “innovating city streets and appearing as an atmosphere of growth and development”. Government support often involves the creation of new regulations in a transparent manner to stimulate investment and growth in innovative and transformative industries and technologies.

Infrastructure improvements

Governments can also provide funding for infrastructure improvements and, through industry consultations, they can clarify how data may or may not be shared. For vehicle safety reasons, as well as regulatory compliance, they can work with the CAV ecosystem to determine IT security standards. By working together, they may also be able to determine how to establish other roadmaps that may define how to safely introduce and operate connected and autonomous vehicles on our streets.

User-centricity

To meet consumer and safety needs, Campion says there is a need for user-centricity because it’s only worth working to develop CAVs with a view someone will eventually pay for the vehicles – whether by purchasing them or hiring them in an Uber-like way.  However, he adds: “It doesn’t help us if we provide technology in a way that causes gridlock, and so there is a need to balance the requirements of the individual with the societal impacts.”

He finds that there is often a technology push, which is about how to make user adopt CAVs. Any attempts to force people to use the vehicles will fail. So, there is a need to focus on what will motivate users to embrace them, by addressing their needs and their concerns. This includes providing features that add value to the connected and autonomous vehicle experience.

Much of this can only be delivered by the automakers.  Yet there is technology available at the moment that permits cars to receive a signal from traffic lights, and which permits them to decelerate or accelerate as they approach them – depending on whether the lights have changed to red or green. “This saves up to 5% of fuel”, he explains before commenting: “This technology only works if the traffic light has been upgraded to send and receive the signal. This means that local government would need to upgrade the traffic lights – perhaps, for example, for someone with a high-end vehicle.”

Singh illustrates why trust will play a key factor in the adoption of CAVs: “With my Tesla I can figure out when I trust the vehicle. Do I have auto-lane changing or not, or do I want to be in control of that action? There is also a need to avoid adding too much flexibility to users. Every time I take over the vehicle, Tesla can see that I have interfered. For certain actions, I’m taking over. So, there is a need for drivers to be able to take full control. For full autonomy I’d advocate the same options, being able to turn off and on certain features.”

He stresses that users you should be able to trust their vehicles 100% when they are in autonomous mode, and so the stimulation of consumer confidence over time is a key factor that needs to be borne in mind from a development perspective.  He therefore advises: “Add flexibility as you grow, and then remove them as users gain more trust in autonomy. At which point I will not need to keep them. With increased adoption I can see how users are interacting with the vehicle, and I may see I no longer need those options.”

Collaborative learning

Campion concludes that it is vital for the different consortia around the world to learn from each other because everyone is at the early state of understanding the benefits of connected and autonomous vehicle technologies. “We should understand the risks and opportunities, rather than the competitive advantages of one particular approach over another,” he suggests.

This process begins by agreeing the ground rules, national and internationally to permit companies to collaborate and to compete with each other. To this end, TRL worked with the European Union on the General Safety Regulations for ADAS, automatic braking and lane-keeping. This is because TRL has a keen interest in “understanding how technologies can play nice with other vehicles and the environment around them. Regulation needs to set a minimum set of standards.” With government support, data modeling and live test data, consortia can help to create them, driving the way forward to connect and autonomous vehicle future on our city roads, and may be even in rural areas.