CONNECTED & AUTONOMOUS VEHICLES
To dare is to do
Max Zadow presents highly compelling evidence to suggest that the UK has justifiable claim to be at the very forefront of the Connected and Autonomous Vehicle movement
“Nobody knows anything... Not one person in the entire… field knows for a certainty what's going to work. Every time out it's a guess and, if you're lucky, an educated one.”
This was screenwriter William Goldman talking about the Motion Picture industry, but it applies just as well to the Connected and Autonomous Vehicles industry. That, counter-intuitively, is actually great news for UK companies.
Tens of billions of dollars worldwide has been invested in developing Connected and Autonomous Vehicles. How much is always an estimate, but the Brookings Institute claimed US$80 billion between 2014 and 2017. In 2017, Chinese tech multinational Baidu set up a three-year US$1.5 billion fund to support development in its Apollo Platform, including investing in global partners. This bought Baidu high level partnerships with Volvo, Ford, Bosch, Nvidia, Intel and Microsoft. Each of these companies is also spending eye-watering amounts on their own internal R&D.
In this sector, where the largest companies in the world are spending billions and working together to achieve their aims, what chance does the UK industry have? The British government itself is fully committed to developing this sector and says it believes that a flexible regulatory environment and dynamic tech SME ecosystem delivers a competitive advantage.
The Government estimates that CAV will be worth £52 billion for the UK by 2035 and in March 2019 announced that on top of previous investments, it would be creating a process to support advanced trials of automated vehicles. That said, investment allocated by the UK Government has been £250 million, which is a fair sum to find in a country living under austerity, but is dwarfed by what is being spent in wider international partnerships.
It is going to be up to UK industry to take this modest stake and turn it into technology that will secure us a seat at the global table in years to come. How? Firstly, nobody knows anything. All those multinationals have been talking about stages in the rollout of CAVs as if they were stops on a roadmap for a well-used highway. This is just an educated guess, if we are lucky. All the strategy that the industry has built on this framework is guesswork based on an estimation. There will be mistakes made, flaws in the logic. In these gaps the UK is well placed to grow its share of the market.
“All the strategy that the industry has built on this framework is guesswork based on an estimation. There will be mistakes made, flaws in the logic. In these gaps the UK is well placed to grow its share of the market”
Estimates of when the next stages of autonomy will occur keep changing. As of the time of writing, Tesla’s Elon Musk is saying Level 4 will be sooner than predicted after years of putting it further in the future, while Ford have revised their plans to remove some of the optimism, which is somewhat ironic, as Ford were bragging before this they would leap straight past Level 3. This was because Level 3 requires you to stop being a passenger and become a driver in an emergency. In tests, most humans were not up to such a rapid change of attention and role, especially in a crisis. Answer? Skip Level 3, as it is difficult.
As a result, this strategy of going straight to Level 4 becomes a statement full of industry machismo. Is it possible? Level 3 is where all the annoying real-world niggles, bugs and sordid negotiations naturally lie. “I think it (Level 3) allows us to provide autonomy in step functions to get people used to it. Acceptance is going to be a big deal; this is totally a different way to move,” says Ford’s Marcy Klevorn, a couple of years after another Ford spokesperson just as confidently announced the opposite.
“CAVs will be bought and used by humans who have complex psychological reasons for doing everything and who have emotional responses to vehicle ownership”
CAVs will be bought and used by humans who have complex psychological reasons for doing everything and who have emotional responses to vehicle ownership (as every car advert tries to take advantage of). Humans have an evolved society that has built up over thousands of years with competing political, economic and philosophical approaches to how every collection of a few dozen streets and houses should be governed. These roads are themselves built on an almost chaotic framework. Roads bend to meet rivers that dried up hundreds of years ago and are a width to accommodate horse drawn carriages (at least in Europe).
This description of our world may seem comical in its simplistic summary. At the same time, many from this industry have attended conferences where visions of how to rip-out and re-build whole sections of cities to fit our Autonomous Vehicle future are seen as an engineering challenge alone. Alterations to every street are going to have to be explained, negotiated and adapted to reality. That is just the impact on roads.
There are people involved too. One of the first industries that CAVs will disrupt is the Taxi sector. Companies like the UK’s FiveAi and Google’s Waymo project are seeing providing taxi services as an initial use case. In fact, most companies in the CAV space take this for granted. It is not often that you hear mentioned the 472,000 taxi drivers who will lose their jobs in the UK alone. They have families that their labour pays for, small shops they spend their profits in and communities who know them. This represents a source of political resistance to the idea of CAVs and a resulting reluctance for the authorities who control the roads to grant access.
“In Liverpool we sparked an anti-5G protest movement, which saw 5G as a threat equivalent to vaccinations, chemtrails and autonomous vehicles. This kind of populist reaction based on conspiracy theory, fear of change and scientific illiteracy could have been predicted but wasn’t”
It is in this penumbral area where CAVs still need drivers, this Level 3, that these political arguments will be had. In Liverpool, we are part of a consortium that has installed a 5G network as one of six of the Department of Digital, Culture, Media and Sport’s 5G Trial and Testbeds. Liverpool City Council joined the consortium knowing the members. They were satisfied the aim of the project was to benefit residents living in the community in which we installed the 5G network.
Even so, we sparked an anti-5G protest movement, which saw 5G as a threat equivalent to vaccinations, chemtrails and autonomous vehicles. This kind of populist reaction based on conspiracy theory, fear of change and scientific illiteracy could have been predicted but wasn’t.
For this 5G trial, our company provided a “digital twin” tool that enabled the efficient line-of-sight positioning of mesh network nodes, in cramped terraced streets rigged for the needs and aesthetic enjoyment of a bustling community. Deploying a WiGig network had never been done on this scale before. The digital twin enabled us test our technology digitally, overcoming unforeseen technical and geographical problems with the network before committing it to the actual street. Why did we have a digital twin available? We had already created one on which to train and verify CAVs (with funding from the Centre for Connected and Autonomous Vehicles, CCAV).
Our digital twin has some advantages over the many others. One of which is a more accurate representation by which Lidar and Radar operates in environments, mostly thanks to our Partnership with Liverpool University and particularly Professor Jason Ralph. Prof Ralph has built up an enviable reputation in the defence industry, including for his work on Lidar and Radar and has been very generous with sharing it for our CAV project. Another company that draws on expertise from military applications of technology is Mobileye. Lidar and Radar is expensive and is confused by rain. Cameras combined with machine learning is the approach that is making Elon Musk feel strangely confident. Sensor setups costing tens of thousands of dollars a test vehicle is not sustainable in a commercial setting. Who knows what will win? Ask Quanergy with their early partnership with Mercedes, innovative Lidar for CAV technology, US$160 million investment and severe recent troubles.
Digital twins are useful, but no map is currently equivalent to the territory so there will always need to be repeated physical trials for the unpredictable to occur and be allowed for. In both these areas the UK has advantages. As well as the Centre for Connected and Autonomous Vehicles focusing sensibly on digital training and verification, the UK’s historically strong computer gaming sector means we have the skill base to excel. Our government, both local and national, is invested in providing testing environments for real-world learning. It is especially useful that these environments are ones in old-fashioned Britain, far from the warm, dry, straight highways of Miami, Texas and California where lack of rain and weird corners provide less challenge to Lidar, Radar and recognition systems from passive cameras.
 FT ibid
Max Zadow is co-founder of Future Coders and collaborator on the ALEAD project on which CGA Simulation are the lead.