Driverless cars; the benefits and disadvantages

self driving car

Written by Josephine Young, Lead Consultant for AI at Methods

Recently the UK Government announced that driverless cars could be on British roads by the end of this year. That’s a fairly incredible announcement.

This is because the UK Government plans to remove the requirement of driverless cars to have a safety human driver in the vehicle, and so the testing of fully driverless cars on public roads can begin. There have been small-scale trials over the past couple of years in places like Coventry, Bristol and Greenwich in London – working towards the end goal of introducing driverless cars by 2021. While there are many unanswered questions (and ongoing debate) about the technological, legal and road safety protections required to make this all happen – the ambition to introduce this technology as soon as possible is fairly clear.

Here at Methods we’re doing a lot of thinking about the impact of emerging technologies– like Artificial Intelligence and Blockchain – and how they can both positively and negatively impact society.

Personally, as someone who lives in a busy, densely populated, expensive and dirty city (um, London) I can’t wait for driverless cars. I think the social impact is going to be overwhelmingly positive in the medium to long term, but I can also see how terribly wrong it could go in the short-term if the introduction of driver-less vehicles onto driver-full roads isn’t managed properly. The impact for regional and rural areas and industries like logistics will be very different, and I won’t go into them here.

So, let me wax lyrical about the medium to long term benefits of driverless cars to an urbanised environment like London… (I’ll get to the negative stuff later).

Medium to long term benefits

All driverless cars will likely be electric vehicles: This means as more of these vehicles take to the roads and replace existing vehicles, the air we breathe will be cleaner. Considering that air pollution in London is criminally high (and I am stating a fact) the need for electric vehicles on our roads is obvious.

Parking might become a nightmare of the past: Currently a car needs to be driven by a human who cannot (and should not) be driving 24/7, which means there’s a large percentage of time that cars are left idle (occupying car spaces). However, theoretically driverless cars won’t ever need to park (unless they are charging) and so each vehicle will be able to complete more trips per hour, and thus reduce the overall volume of vehicles required to move our great city around.

Improved levels of safety for women and people who are vulnerable when they travel by cabs or rideshare services: Basically, if it’s important for people’s safety and independence to be in the car on their own when they travel, then a driverless car can provide that. And often, for women, people with disabilities, and people in the LGBTQI community, travelling through public space is an unsafe act for them. As an indication of how much of a problem this currently is, TfL’s stripping of Uber’s license to operate in London in 2017 was in part due to police figures showing that an allegation of rape or sexual assault is made against Uber drivers every eleven days.

Roads that have a high volume of driverless cars on them will be safer: This is because the reaction times of driverless cars will be faster than driver-full cars and the layers of car sensors and software will constantly be working to minimise any potential dangers. In addition, if each car is on the same digital driverless car network then the vehicles will actually be able to speak to one another and ‘know’ where and what the other driverless cars are doing. Driver-full cars could tap into this network by carrying their own sensors, making sure that their vehicle is partially integrated as well.

Building on the idea of a driverless car network; with greater integration of vehicles there would be an opportunity to link them in with road infrastructure like traffic lights. Imagine if all the cars in the network were able to move around the city at speeds and on routes that minimised bottlenecks, traffic jams and getting caught by every single red light on the way to Tesco.

And lastly, imagine if we integrated public transport (including Santander bikes) and emergency services vehicles into the driverless car network. Driverless cars could pull over for emergency service vehicles, knowing when they are on their way. Areas well-served by bus routes could have a lower priority for driverless cars so that these roads experience less congestion (and faster bus trips!) and prioritised driverless vehicle capacity for areas less well-connected to the existing public transport infrastructure.  And imagine if you could pay for your driverless car journey with your Oyster card…

This leads us to a very big, structural question: who will own and run the driverless car network?

What could go very wrong in short term

Answering the question around who will own and run the driverless car network requires public debate and thinking around:

1. Figuring out what we mean by ‘safe’ driverless cars and who’s responsible when something goes wrong.

2. Taking a social good and social justice approach to the role of driverless cars in our society.

And to be honest, what follows are a series of unanswered questions floating around in my head (whereas in the previous section you got a glimpse into my driverless car-related daydreaming).

1. For the first point on ‘safe’ driverless cars, there are a number of issues that have popped up long before driverless cars began testing on public roads. Firstly, what is the minimum level of safety we require from driverless cars? Do they need to outperform humans on roads in order to qualify? Is even one death or injury, as the result of a driverless car, one too many? If safety isn’t made clear and rigorously enforced, then public trust in driverless cars will evaporate and there will be many ways in which society will resist their introduction.

Secondly, how will driverless cars respond to deliberately reckless driving from the humans also occupying the roads? And what needs to be put in place to ensure people don’t take driverless cars as implied permission to relax their own diligence when driving on the road? A lot of people could get hurt in the short term if the overall safety of roads is compromised due to human-led reckless driving, thinking that it might be funny to try to ‘mess with’ the driverless vehicle or that the driverless vehicles are the only agents responsible for collective road safety.

Thirdly, how will we make sure the driverless cars and the network they rely on aren’t hacked and potentially vulnerable to control by someone other than the passenger? Imagine if the foreign-owned misinformation bots turned from undermining Twitter in the lead up to a national election to gaze wickedly at London’s morning peak hour traffic.

Finally, which human will be responsible for the behaviour of the driverless vehicle on the road? The owner? The passenger? The company? The car itself? Once this fundamental question can be answered it will be possible to imagine the construction, regulation and use of a driverless car network.

2. Moving on to the second issue: building a social good and social justice approach to the role of driverless cars in our society. If we get this right then we can expect the medium to longer term benefits to materialise. If we get this wrong, however, we’ll have missed a massive opportunity and risk the benefits of driverless cars being enjoyed by a small (privileged) part of society.

So – don’t just hand it over to a tech company. While they are the organisations producing this truly impressive technology, they don’t necessarily have the expertise or capability to run and manage car fleets or road infrastructure. They will also tend towards building systems and ways of running these vehicles that focus on maximising customer satisfaction and their shareholders value (likely predicated on intense surveillance data collection). This focus (while understandable) misses out on maximising community safety, redistribution of benefit across all layers of society, and designing to minimise social and ethical issues. Any introduction of driverless vehicles will of course need to see financial benefits flow back to the companies that invented the technology, but that doesn’t necessarily mean they’re the most qualified to lead the deployment of it across society.

And, building on the previous point, there needs to be an organisation that takes responsibility for supporting those who will experience displacement as a result of driverless cars (see Black Cabs, Uber drivers, lorry drivers). Can we imagine the large tech companies doing this? Probably not, no. Can we imagine a government body going this? Probably, yes. Which leads us to further questions of how this support will be funded, to what end and specifically by whom.

Importantly, an introduction of driverless cars needs to be underpinned by the following principle: driverless cars must increase the freedom of movement for everyone in society. Crucially, how can people with the least means at their disposal, or living with the least access to affordable transport options, have their freedom of movement around London dramatically improved? Any system designed to fulfil this principle will also see the benefits flow to people who don’t currently experience these difficulties. Designing for those least served right now, ensures access for everyone in society.

Personally, I think the driverless car network needs to be government owned (by an organisation with a trusted record in this space – e.g. TfL), set up in a way that it is affordable and accessible to everyone, solves issues across the transport network and takes care of those who will be negatively impacted by this transformation.

In short, give me a driverless car I can order through TfL and pay for with my Oyster card, that helps ambulances get to hospital faster, makes every journey on British roads a safe one, and means my neighbours don’t ever have to choose between transport and heating. That’s a driverless future I can get excited about.


Originally posted here.

More thought leadership

Comments are closed.