Autonomous Vehicles: Part 1
Posted June 29, 2018 by Roseanne
Hear what James McKemey, Head of Our Insights Team, has to say about Autonmous Vehicles. Another very hot topic around the future of transport...
At Pod Point, we’re big fans of Tony Seba’s Clean Disruption book. If you have a spare hour you can also watch his presentation on the same subject here.
Tony makes the case that three things are coalescing to utterly disrupt several industries:
1. Electrification of transport
It’s better and will soon be cheaper to purchase as well as use.
2. Solar’s imminent emergence as a dominant power source
Tony shows that solar power will soon be cheaper per kWh than the cost of transmission of conventional power (though of course, it makes sense quicker in California than in Dornoch!).
3. Autonomous vehicles (AVs)
The removal of the paid for driver (and switch to electric motoring) making ride sharing so incredibly inexpensive that it will encourage us all to ditch our cars and just hail rides. This will have profound effects on both the automotive industry, but Tony also predicts changes will be as great on the urban landscape, as we can start to do without so much space for parking.
Tony says you could fit 3 San Francisco’s into LA parking alone!
Now, just because we’re big fans doesn’t mean we agree with everything Tony says wholesale. And AVs is one where we’re not quite as gung-ho as Mr Seba. Let’s take a look...
Easy one. Very electric. Wholly electric. We are 100% aligned with Tony here. Electrification is coming and it will drive the costs out of all motoring, along with mitigating its environmental impact. AVs won’t succeed unless they are A Electric Vs (AEVs). This will also have impacts on charging infrastructure requirement.
When will AEVs be ready?
Now this is a big question and there are differing opinions on it, which I’ll swiftly review.
a) Decades away, if at all
There are those who think AEVs aren’t that valuable. There is a loss of utility of the driver e.g. helping you with luggage. There is the economic impact of the demise of the vast driving industry. There is the reluctance of drivers to purchase AEVs or do away with driving altogether. There is the technical challenge (things like the “Holborn Problem”). There is the regulatory challenge of getting them permitted to operate in the wild. There have been high profile crashes of prototype AVs. There are concerns over cleanliness. All fair concerns.
b) Soon! Like it’s kind of here already and will only grow
Probably the leading AV firm is Google’s Waymo. They have been safely operating genuine driverless cars on the streets of Arizona since 2017. This alone disproves some of the naysayers, but it is in somewhat controlled conditions. You have been able to watch a Tesla Model X drive an employee to work, drop him at the door and then go park since mid 2016, and Tesla says every car they’ve made since November 2016 has all the necessary hardware to operate as an AEV (though they now say they’ll have to swap to a more powerful onboard processor). Many, if not all, OEMs have significant AV programmes, BMW’s recently covered by the BBC.
EXCLUSIVE: You are looking at a level 5 autonomous vehicle - the holy grail in self-driving cars - an auto that can, in theory, drive anywhere at any speed, without human intervention. Later I’ll be taking it for a spin. Or should that be it’ll be taking ME for a spin?! pic.twitter.com/d8yISP9GEt— Victoria Fritz (@VFritzNews) 11 April 2018
Alright James, get off the fence…
I think fully AVs will be a common sight on the road within a decade, but probably not until circa 2025. The business model challenge is pretty easy, lose the driver, save a bunch of cost. It’s the technical and regulatory challenges that are still to solve.
- Technical challenge
It is bewilderingly complex to make an AV that performs the same functions as humans. You have certain advantages (perfect 360 vision, radar etc), but the nuances in driving are truly immense. Think about all of the information you are processing, the subtle human interactions, all those hazards you learn to perceive. And I do not imagine there will be a hard line where one-day humans drive and the next the robots do, so the robots need to start in the irrational human driver world first.
However, though the challenge is big, the sheer might tackling the challenge is, or soon will be, even bigger, i.e. artificial intelligence and machine learning. Drive 10,000 miles a year from turning 17 until they pry you out the car at 85, you will gain 650,000 miles of driving experience. Stick censors in every vehicle you make that maps the environment and records driving behavioural data, take that vast data set, apply machine learning and use it to shape your AI driver - you can get that level of knowledge multiple times per day.
- Regulatory challenge
This is arguably harder. Governments… erm… do not move with the speed of tech giants. Getting this approved will be arduous. But it needn’t be impossible. And several governments have shown some political will to move it along. In particular, the UK government have twinned EVs and AVs in a preparatory bill called the “Automated and Electric Vehicles Bill”, seeking to give the government powers to speed the arrival of EVs and AVs, which is a clear statement of political will to get this accomplished.
That said, it will happen territory by territory and there’s no doubt some will take a dim view on the threat to jobs in the driving economy.
Stay tuned for part 2 of James McKemey's thoughts on 'how big a deal AEVs will be'!