Why We Should Go Driverless

Why we should go driverless with cars

So let's talk about why we should go with driverless cars - whether you like it or not, it's coming! The BBC reports that this may be coming as early as spring 2021 - so lets have a look at what this means.

The Economic Advantages of Driverless Cars

So, the economic advantages may seem obvious, and this is the main reason the UK should be leading the way to allowing driverless cars, especially if we want to make a big dent in the economic hole left by the Coronavirus.  If I just look at my own personal circumstances, I drive around 16,000 miles a year at an average speed of around 40 mph (I’ve done a bit of rounding to make the maths easier), that’s around 400 hours in the driving seat or an extra 10 working weeks of recoverable time, that you can use for work, play or family time.  That’s just the tip of the iceberg though, a driverless lorry for example could run for 3 times as many hours at half the price, saving millions on distribution costs.

From an individuals point of view, second car ownership would decline as it would become significantly cheaper to get a driverless taxi, the most expensive part of a taxi currently being the driver.  You would have an app to request a collection and state destination and arrival time, and a car would come and collect you with discounts offered for shared occupancy, as the service gained in popularity, costs would plummet as the per mile cost for a car to do 300,000 miles over 4 years with multiple occupants is significantly cheaper than the costs for an individual to do the 12,000 miles or so a year that the average driver does.  The decline in second car ownership will reduce the need for as many parking spaces at shops, offices and homes and the increase in shared occupancy will reduce the traffic on the roads and reduce the need for new road schemes for a number of years to come.  This, as you can see, will save industry, government and the individual significant amounts of money and getting there first would put the country at a significant advantage over the rest of the world.

It’s not all good news on the economic front though, driverless delivery vans would lower costs further for the online delivery industry and in so doing putting another nail in the High Street coffin, and the 100,000’s of driving jobs up and down the country would be lost creating a bit of an unemployment issue (however, this extra labour availability is also an opportunity but this article is about driverless cars so I’m going to skip over that) and there would be a bit of a slump in the second hand car market.

The Environmental Advantages

Hand in hand with a reduction in car ownership and shared trips is a corresponding benefit to the environment, both in terms of a reduction in number of trips and the lighter traffic making driving more efficient.  Additionally, the better utilisation of vehicles will have further environmental benefits both in terms of production, delivery and disposal.  Having a low cost, on demand taxi service could also increase the use of (driverless*) trains for longer journeys, especially for lone travellers, where a driverless taxi will take care of the transfers at either end.

_{*I fought the urge to raise the subject of whether or not a train needs a driver … I lost that argument!}

You Want to be in Control

So, you want to be in control, but why is this? Surely, with today’s traffic, it’s not for the enjoyment of it?! The main reason that people tend to give is that they do not trust the technology. I have driven a driverless capable car for a number of months now and I can tell you that the technology is far from it’s infancy. But would I trust it to drive myself and my loved ones around without me being fully alert and ready to take control at all times? Well “yes” … and “no”.

Yes, because data has now been collected over millions and millions of miles, and the driverless technology is now better than the average driver. Furthermore, the insurance companies know exactly what the risk is for a driverless car based on the millions of miles of data, and they can tell you which roads they are better on. Yes also because my experience is that the margin of safety that they build into the system is far higher than the margin of safety that most human drivers operate within, in fact when I’m in control of the wheel, the car is much more panicky of potential hazards than the most nervous of passengers.

However, I also said “No”, and for 2 reasons.

The first is that whilst driverless systems are very good on some roads, motorways, A-roads, B-roads and many urban streets, they are less reliable on country roads and other roads where the environment is less well defined. I don’t see this as a big problem though, there is enough data now to be able to work out where a driverless car would be safe and where it would not and we have the technology to only allow a driverless car on the safe roads with councils given the task to ensure roads are maintained to a standard to allow access to most of the population. 

The second reason is not a technology argument, but an ethical argument. If I am in control of a car and the unthinkable happens then, as a driver, my instinct would be to do everything in my power to ensure that I protect myself and my loved ones from serious injury. However there could be a situation where the least number of fatalities would be to drive my car into a wall, and I’m not sure I would be comfortable having a computer making that decision for me and I would want to know that my car was looking after me first and foremost.  This brings me on nicely to the next point …

Who Can We Blame if a Driverless Car Has an Accident?

Now, this is not a big issue for minor shunts, however if someone ends up dying, then current thinking is that we need someone to blame, someone to take to court, and potentially someone to send to prison. So how does this relate to the technology behind driverless cars?

Well, I’ve already touched on the insurance implications above. Currently, your risk is based on your past history, your age, where you live, your occupation etc., this is all very much an approximation and gives rise to significant variations in insurance premiums, however with driverless technology, the risk can be calculated much more precisely for a given road and conditions. From an insurance point of view you can specify what roads and conditions would be covered and you could technically prevent the car from operating outside those parameters.  So, the risk in insurance is easier to define than for an individual human … but what if someone dies?

This is where regulation needs to come into it, providing that the driverless system has undergone significant testing, and updates are significantly tested and do enough test miles prior to release etc., and that this is fully regulated, then that needs to be acceptable.  Don’t get me wrong, the likelihood is that someone will die eventually and that will be unfortunate. As drivers, it is a risk that we take every time we get in a car, and it is a risk that millions of people in the UK are willing to take every single day; the fact that we could be involved in an accident with someone, statistically, is a much higher risk than a driverless car.

There are thousands who die each year at the hands of human drivers, many are accidents, some are reckless, some are down to tiredness or substance abuse, some are down to mechanical failure or health issues, some are down to incompetence. So who’s fault are those accidents, is it the manufacturer who made the car, the driving inspector who passed the driver or the government who issued the license, all of whom did not do a thorough enough due diligence to prevent those accidents from happening.  In the case of driverless cars however, we have the data now to know that, on certain roads, the driverless car is much safer than the average human driver and it gets better with every mile driven, so why don’t we allow this and start saving lives? Worrying about who to blame for a single death is probably costing many lives, putting the legislation and regulation in place to allow the use of driverless cars is, in the words of Thanos, inevitable!

The UK should Lead the way

With all of the benefits, the UK should really look at how we can adopt this technology to lead the way. We already have an economy that is hampered by the quantity of vehicles on the road vs the amount of road available. Furthermore, a growth in technology has always coincided with success, and being resistant to the change will only hamper the country's economy. 

"Self Driving" Software - AI

So why is the MD of a leading business software solution provider talking about driverless cars. Well, the parallels with the modern business software solution are considerable.

Much of the same technology behind driverless cars is starting to be used in the world of business software, Artificial Intelligence (AI).  Whilst I have not mentioned Artificial Intelligence in the narrative above, it is Artificial Intelligence that is driving the driverless car, and it is Artificial Intelligence that is driving the benefits of the modern business software solutions.Microsoft, with their Dynamics 365 Solutions, Power Suite, Microsoft 365 and Azure based artificial intelligence solutions is certainly leading the way in this development.  There are of course other AI solutions … Alexa, Google, Watson … but who else has the range of business applications being developed in parallel to make use of Artificial Intelligence? Its seems like it's just Microsoft.

With AI solutions to calculate stock levels, answer queries, determine the sentiment in the market place, process orders from an email, convert speech to text, language translations, manage actions, manage production, quality control, recognition of objects or people, and a whole host of other uses, the economic advantages of adopting AI technology is clear.  Even if you are not realising any direct cost savings with the adoption of AI, the value add with improved service should put you a step above your competition.

From an environmental point of view, making use of solutions to improve distribution efficiency, reduce machine running times, reduce incorrect deliveries and resolve issues remotely without the need for a return or onsite support will help reduce the carbon footprint.

And then there is the control issue. Anyone who has grown a business from a small company to 30 or more people, knows there comes a time when you need to relinquish control, to put your trust in others to do a job that is good as (or better than) the job you want them to do. At the moment, most people put their faith in other people, and with the help of good processes this usually gets them to where they want to be, but there are still mistakes along the way, things that go wrong, people forgetting what should be done. With Artificial Intelligence, yes, there can be mistakes along the way while AI learns a process, but like with an individual, you correct AI and it learns from this. The main difference is that when AI learns something, it doesn’t usually forget, doesn’t take holidays and you don’t have to train a temp.

So, who is to blame if something goes wrong?  Well, in business you usually have to take it on the chin, it is the company’s fault, no matter who or what in the organisation makes the mistake, and responsibility ultimately resides on the shoulders of the directors.  So why would a director put their faith in an AI solution. Well, for a non-critical solution, the benefits above should be a enough. But for a critical solution, a life or death situation, then would the AI solution not do a better job than a person?  If you are not willing to take that risk, then the belt and braces approach would be to run both.

So, just like the UK should lead the way in adopting driverless cars, shouldn’t your business lead the way in adopting driverless business solutions and gain the advantage over the competition?