Will half the US really be 20 years behind the driverless revolution?
We are now accustomed to seeing celebratory back-slapping PR and press coverage as driverless vehicles entered the public conscience around the world in 2017, but given the gravitational pull of the world’s autonomous industries towards places like Silicon Valley, I started to think about what a real tour of the USA would look like if I were to deliberately avoid its two greatest epicentres of automotive technology activity, California and Michigan.
Who is doing what, and where, what are the impacts of those businesses and what are the other side effects of these industrial, commercial and technological endeavours?
To help my research, I dove into our website reporting, and explored two years’ worth of data, finding visitors from some 5,000 organisations in more than 75 countries, of which roughly one quarter was in the USA. For the sake of clarity, I have used the organisations main headquarters as the reference point, so while there will be jobs and factories in every state, I needed to restrict the data points fairly and reduce duplication. From this point, I felt that I had enough geographic information to start a little tour, from my desk in rural England, and make some reasonably confident conclusions about what I found.
The first conclusion I reached was that, between small populations and even smaller population density, more than half of the States in the US will be playing catch-up, long after autonomous vehicles are seen in major cities and even longer after they are seen working experimentally in many more.
Missouri, South Carolina, Kansas, Nevada, Utah, Idaho, Louisiana, New Hampshire, New Mexico, Maine, Nebraska, Oklahoma, Vermont, Arkansas, Kentucky, West Virginia, Wyoming, Alaska, Delaware, Hawaii, Mississippi, Montana, North Dakota and South Dakota – this is the list of states that are really not the right place to be if your first love is autonomous vehicles.
Watch AutoSens TV ~ (Click to subscribe)
► Artificial Intelligence – The Future of Transportation?
► Learnings from creating a training dataset for autonomous driving
► Intelligent sensing requires a flexible vision processing architecture
But why? With modest populations, most also have only a small handful of companies investing in autonomous vehicle technologies or research, some not even able to boast even one university or college investing in important future skills.
California, on the other hand, unsurprisingly at the top of the list, boasts more than 25 colleges and universities with an interest in the field – that’s more than the organisations of all types in the bottom 15 states put together.
A Downward Spiral
Predominantly relying on non-technical industries such as agriculture, when it does arrive in these outlying states, autonomy is likely to come first to cities and then to industries where comparatively repetitive low-skill but high-volume jobs are mostly likely to be lost through automation, such as road freight, mining, farming and forestry. These are hardly surprising statements, and already well covered by politicians, strategists and researchers.
But the conversation on autonomy in the rural environment is complex and might be better served by simply highlighting the need for improved road safety, rather than validating a commercial justification for driverless ride-hailing services, which currently seems to be the largest volume catalyst in this new rush for transport technology deployment.
Could these state economies be the last to benefit from the dramatic improvement in road safety that increasing levels of safety systems and autonomy will bring? You might laugh at the idea that safer cars are the preserve of the wealthy, there is also a distinctly vicious cycle effect to having more dangerous cars on the road. Higher levels of road accidents.
Road accidents mean high costs to the healthcare system, insurers, drivers and of course – has a huge negative knock-on effect on the local economy. People that are severely injured are unlikely to be contributing a great deal to their employer, and worse still, people killed in road accidents are even less economically productive.
Of the top 10 states with most activity in the autonomous vehicle sector, we find only 1 state (Texas) which is in the poorest performing half of the US when you look at road accident rates. In those 25 states, which have very little activity in autonomy, we can count 16 (a surprising 64%) that are in the worst half, of all States, in terms of rates of fatal road accidents.
Even if we took an average road accident rate from the top 10 states, vs the bottom 25, you are 65% more likely to be involved in a fatal road accident in those I’ve listed above. That’s on average – that’s not even comparing the best state with the worst state.
There’s irony then, that states whose economies are most damaged by road accidents are investing the least, encouraging the least and understanding far fewer benefits from the coming changes brought by autonomy.
Autonomy – Who’s going to be first?
That’s no longer the competition. Being first is not relevant, but being the best might help. Many countries will see the benefits sooner thanks to robust education eco-systems, supportive investment from private and public sectors, and a ready stream of entrepreneurs and technology buyers. The USA will – at least in part – lead the way, followed by the major automaking and technically adept nations such as Germany, Japan, China, UK, France, South Korea.
Could we make some broad assumptions that the global roll-out of driverless services could be played out in miniature inside the USA; the world in microcosm?
From the remote and frozen, to the heavily forested, the mountainous to the arid – America’s rural extremities and geographic diversity mirror the conditions we find in the majority of other countries around the world. While language, road culture, environment and education might change, could we really assume that Florida really be economically as impactful as the whole of Brazil?
If we are confident to draw statistical equivalents, it’s important we also recognise a correlation, not causality. There will be a ratio between investment and safety improvements, but there are so many other factors in play (not least legislative, cultural, industrial and societal) that picking a state from the list and comparing it with a country would never stand up to rigorous analysis. After all, if autonomous vehicles are proven worthwhile at Level 5 in California, they will inevitably be deployed quickly across all US states – including those that have not invested – but the remorseless expanse of the huge and sparsely populated countries of Russia, Brazil and Australia are unlikely to see the benefit as quickly as from, say, a digital product.
- Agenda announced for AutoSens Detroit 2018 ~ 23 February 18
- Why our greatest success is a disappointing reflection of the automotive industry ~ 23 February 18
- Security and Safety challenges in Deep Learning for Autonomous Driving ~ 20 February 18
Sure, it took 20 years for TVs to be found in 50 million homes, but Facebook did it in 3 years. Apples and oranges… or more accurately, apps and TVs. But locally made TVs that don’t have the latest feature don’t have a negative effect on your country’s economy – unlike the safest cars with the latest ADAS.
So, investment in this technology could correlate to reduction in that road safety impact I mentioned earlier. The more you invest now – as a government, be it national or state – the fewer negatives you will see. Fewer road deaths: a healthier economy. Irrespective of the level of technology deployed, ADAS and autonomous vehicles are still a physical product, which must be physically created, physically moved, maintained and paid for with real money. They cannot be downloaded, they must – at least for the next decade or so – be designed, made and handled, until they are ubiquitous.
That simple fact alone will slow their relentless roll-out across our planet.. and as the frenetic pace, of investment and technological progress in ADAS and autonomous vehicle technology consumes tax subsidies, VC investment and insurance against future obsolescence, this limitation – reliance on humans and the challenges of the physical world – will help calm things down. Should it also provide a warning to those countries and communities not embracing?
Are they really saying “this fight is not for us, we’re happy to lose 3% of our GDP every year”? As an industry, we expect to enter the ‘trough of disillusionment’ whenever a sufficiently new technology is promised. For many, the money is already in the bank – at least in those places enlightened enough to make this a priority, now it’s time to crack on and solve it.
If we solve it in our industrial epicentres today, it’s still going to take 20 years before we have a chance of solving it everywhere else. The problem isn’t going any where, and luckily, neither are we.