vSearch by Date

June 2019 - June 2006

vSearch by Title

According To Me [Autonomous Insurance] - Autonomous Driving



June 2019

A coloured pencil drawing of a three lane motorway from directly above showing a number of close together vehicles of various types

Autonomous Vehicles

Part One - An Introduction

The latest news in driving is that driving is to expire. For us mortals at least. Soon the only thing driving our cars will be the cars themselves. Yes, all over the news we hear of self driving cars. Just use a common search engine to see who is big in this field and you'll instantly get the idea. I suggest the search engine Bing.

Some vehicles already have lane departure systems that bleep at you or shake your steering wheel if you dare to cross the line markings without first advising the car. Many more have cruise control to avoid us having to make the effort to maintain a speed, sometimes enhanced with additional radar control to keep us from accidentally bumping into the vehicle ahead. We have self parking systems to get us into a gap and detailed mapping to get us out of holes. Although to map out all the actual potholes encountered may take another 30 years.

This is all big news and for those that follow my every word across all the social media platforms that I use, yes you two, you will be well aware that I have a great fascination in this sort of technology with the development of autonomous vehicles being of the most interest to me. In fact I have been picking at this subject for a few years now, as I shall demonstrate.

My first ever public comment on any aspect of autonomous driving was made on the Twitter platform back on 1 June 2014 when I posted the thought 'Can't wait for these driverless Google cars. Will make my border drug running business a lot less risky'. And if you think I have blown a cover on an illicit controlled substance operation then you haven’t been following my Twitter stream very carefully.

Then, after reading about a potential development on a Honda Accord car that would use 'radar, cruise control and the ability to follow white line markings whilst steering to effectively allow the car to drive itself' I posted a blog on my web site on 14 June 2006 entitled 'According To Me' [link below]. In this I mused over a potential dispute between various interested parties in the event of a collision of an autonomous vehicle.

I continued within Twitter on 17 June 2014 publishing another tongue-in-cheek tweet writing 'Love the internet technology on new cars. Just emailed my brakes. Now waiting for a reply' with another post a month later on 17 July 2014 wherein I wryly mused 'If spell check gets in the way, in the future will Google produce self driving cats?'

By 3 September 2014 I had more to say on related matters in this field which I literally did within my fifth podcast subtitled Lanserguided [link below]. Please feel free to check out the whole aural experience but if you find the idea of my voice droning on then in essence I raised an idea about potential laser projections on the front of cars to map out a stopping distance ahead of a moving vehicle. Then I considered whether future autonomous cars would actually allow us to get into them or consider driving away completely if sent off to seek a parking space. I also predicted a simpler future driving test. Plus I concluded that the take up of autonomous technology would be inevitable. I did offer a caveat that despite all the promise of automation there will always be human skills needed to maintain and service broken vehicles.

In addition to this early sporadic public commentary on the subject I had many other thoughts on this developing and fascinating technology but the next tranche of public comments came again on Twitter in a series of tweet posts over a year later on 26 October 2015 as follows:

If I bought a driverless car and sent it to park while I was at work, what's to stop it starting it's own taxi service?
Of course the wealthy already have driverless cars. Or as they call them, chauffeurs
Personally I'm waiting for the first fight between driverless cars over a parking spot. That may sort out the Android vs iOS argument
I bought a driverless car last year. It read roads, maps, the Internet & communicated. It went straight to the High Court & claimed freedom

Using a different medium, this year I tried some stand up comedy and for one performance I wrote a routine about driverless technology which I performed at The Studio within The Point at Eastleigh, Hampshire on 20 February 2019. It was a deliberately light hearted slant on the subject but did cover many interesting points within this field. You can view the performance on YouTube [link below] but I have extracted some of the ideas here for information whilst simultaneously extracting the humour because this is a serious article and nothing even remotely amusing must colour the tone. Ever.

Autonomous cars, driving around with no apparent attention being paid at all. Is that taxi drivers?
How do you operate a driverless car? If you’re wealthy, use voice commands. It lets the chauffeur know where you want to go. For the rest of us you’re no more than a dog. Open the hatch of your Rover, get in and it takes you to your destination
Future driving tests will be so much easier. “Show me your car. Get in. Seatbelt on. Good, that’s a pass”
Cars will be able to communicate to work anything out as a group. At a traffic light on the illumination of green they all move off at the same time and on red they all stop as one
All controlled within parameters of the users choosing a priority mode of travel - Tourist mode, in a hurry or even declaration of an emergency. Enhanced by the cars choosing priority based on types of occupant
On the open road cars will be able to go really quickly with future motorways packed tight full of high speed cars, all talking to each other
Are our current cars going to be scrapped being no longer useful?
Are we are going to have to fit our present cars with similar cameras, lasers, radars and sensors similar to those needed for the autonomous ones?
What if we have super smart self driving robots that can get into our current cars whilst still quickly communicating with the new tranche of driverless vehicles? These to connect with all the other cars, fitted with multiple ‘limbs’ for steering, gear changes, handbrake, wipers, lights etc. Plus being plugged into the car’s on-board computer and fitted with all necessary cameras, lasers, radars and sensors all over to simultaneously look out the front, the back, see the traffic, see every single mirror, check the speed, revs and fuel gauge etc.

And my latest public commentary on the subject came on 27 March 2019 when I tweeted 'If my car camera automatically reads speed limit plates to restrict my progress I may need to tape a photo of a national speed limit sign onto the end of a fishing rod and hang it out the front'.

So below I expand on these thoughts posing a number of ideas, questions and ideas on this subject to summarise my position, as follows:

Autonomous Vehicles - A Transitional Period
Autonomous Vehicles - Issues and Scenarios
Autonomous Vehicles - The Future
Autonomous Vehicles - Interesting Questions and Considerations

Part Two - A Transitional Period

This section will look at my thoughts on the transitional period between full driver control and full autonomous control. But before we proceed too far what is meant by an autonomous vehicle?

According to Wikipedia, which is the best source I can suggest if you wish to know more, automated driving systems were first trialled as early as the 1920s. However autonomous driving as we know it now with greater control by electronic means was worked on in the 1980s but it wasn't until the 2010s and the development of more powerful and cheaper computer systems that modern recognised autonomy tests were being carried out.

For information the most accepted standard of defining autonomous control is from SAE International, an automotive standardisation body, which defines levels of driving automation as follows:

Level 0 - Basic - Automated warnings and momentarily intervention but no sustained vehicle control
Level 1 - Hands on - The driver and the automated system share control. Examples include Cruise Control, Adaptive Course Control, Parking Assistance and Lane Keeping Assist. The driver must be ready to retake full control at any time
Level 2 - Hands off - The automated system takes full control of the vehicle with the driver monitoring and regularly demonstrating this and fully ready to intervene immediately at any time if the automated system fails to respond properly
Level 3 - Eyes off - The driver can safely turn their attention away from the automated driving but be prepared to intervene within some limited time
Level 4 - Mind off - As level 3, but no driver attention is ever required for safety and could sleep or leave the driving seat. The vehicle must be able to safely abort the trip if the driver does not retake control
Level 5 - Steering wheel optional - No human intervention is required at all

There are problems transitioning between these various levels of autonomy, particularly from levels 2 and 3 and major car manufacturer Ford announced in February 2017 that they gave up attempting to develop a level 3 stage opting for working on the safer level 4.

Well, that's the vital but soulless bit out of the way so I can now concentrate on my own related thoughts.

This may surprise some but we already have a great deal of autonomy in our cars, in fact every driver who has ever driven has experienced some form of autonomy because some aspects actually go back to the earliest days of motorised transport.

A photograph of a Benz Patent-Motorwagen taken indoors at a Motor Exhibition
The Benz Patent-Motorwagen, the very first automobile with some partial automation arguably making it technically the first level 1 autonomous vehicle

For example the very first motor car, the Benz Patent-Motorwagen required a considerable amount of driver input control. Speed, braking and steering were all completely operated by the person in the driver's seat. However some of the mechanical operation had been automated, for instance the trembler-coil ignition system. This was technically an automated device to avoid the driver having to manually open and close an ignition switch for each stroke of the engine.

If you feel that I am stretching a definition of autonomy then where do you put your line? On the subject of ignition again is it an autonomous function for a car to adjust advance and retard on a car without electronic ignition? Cars used to have advance and retard for their ignition to be manually selected by the driver dependant on the incline and thus load on the engine. Now it's the car that works all this out and only a retard would dismiss this advance. This function used to be manually operated and is now an automated function, with a great deal further development in the electronic ignition systems of modern vehicles.

Perhaps your 'line' is drawn at automated gearboxes. These are fairly usual nowadays particularly as manufacturers have improved the efficiency of auto gearboxes. I ask, why choose the manual option if not for cost? Nowadays I'd much rather drive virtually any auto than a fiddly manual because I just can't be bothered with all that awkward left leg clutch balancing stuff, particularly in our congested, traffic jammed streets. And if I really want to drop down a gear for added oomph, or preselect a cog for better downhill control I can always flip the flappy paddle.

Now gearboxes have gone beyond just selecting gears based on engine pre-selected power [revs] and speed. The Volkswagon Group DSG type gearboxes are designed to pre-select gears based on assumed future driver requirement and Rolls-Royce provide a gearbox in their latest Phantom VIII model which means it can preselect a gear based on GPS receiver and terrain information drawn from a map system.

And even if you don't accept these gearbox functions as autonomous features then surely many will have driven a car with cruise control fitted, which is a defined level 1 stage of autonomy. And some may even have had a go at successfully operating it.

I personally have experience of driving vehicles to at least level 2 autonomy which are now often fitted with technology to level 3 but restricted in use due to legislation. This level is not uncommon on modern vehicles with adaptive cruise control, lane keeping assist, automated lane changing and automatic braking. Although I accept at present these systems are often unreliable in some circumstances, such as when the road lines fade away or within the confines of supermarket car parks.

I have no experience of driving to level 4 known as 'Mind Off', but I reckon many have. Usually just before they crash!

So, as you can see we are already accepting the partial automation of our driving. The transitional period is already here with us.

So how will we actually ‘drive’ a true self-driving car? What will be the process?

I have already suggested that operating a level 5 autonomous car would be so simple that a future driving test would be almost completely pointless. However, we would need to show some control, after all the car would need to be summoned, opened and told where to go so manufacturers should make these basic steps as simple as possible to enable maximise occupancy use, for occupants such as the old, infirm, children or dogs that wish to go walkies while their owner watches the snooker on TV.

However when in the interior space it won’t be a free-for-all, at least not for now. You will not be able to meander around a vehicle making tea and dancing to Reggae music. The laws of physics still apply so when the vehicle accelerates, brakes and corners you need to be securely strapped in to avoid spilling that tea or turning your Jamaican moves into an impromptu break dance. Whether we get to the stage where accidents are so rare and driving so smooth that full freedom of movement will be allowed in the vehicle is yet to be proved.

So what are the level 5 control systems likely to include? A reasonable guess is that within cities the vehicle may not be where you are when you first summon it. There is a high likelihood of car sharing in the future and possibly the common use of tightly packed, remote storage. If cars can operate autonomously why would they need to remain outside your property getting in the way of the rest of your life? They are more likely to drive themselves off to a charging point or park in a remote car park, possibly automatically stacked several cars high or at least packed in nose to tail with no room between them to open doors. For this reason to get in your car you will first need to summon it and this will in all probability be an application on your mobile device. By then we would have to be accustomed to summoning a ride in good time and if it was a genuine emergency the system will just divert a closer 'common use' vehicle.

Just before the vehicle arrives it will probably be sending out a message for you to let you be aware of when it will get there and once at your location will be linked to your mobile device so that it is open as soon as you are ready. A quick, bluetooth style, electronic 'handshake' between you and the vehicle and it will be ready for instructions on where to go next.

Of course it won't move off until all occupants are secured into a seat, so no racing starts to beat your neighbour to the end of the Close. The seats will then probably be able to determine who is sat there and make suitable adjustments that the occupant has used before, whether it be facing ahead, in a face to face group gathering or prone for a sleepy ride.

One person will then probably state a destination requirement, such as "Hey car, take us to The Dog & Duck, via Barry's place." This will prompt the car to respond such as "You wish to go to Kentucky to buy some milk, please confirm." Or at least it will do if voice command doesn't get any better. But in any case the vehicle will have a tablet device slotted in somewhere so that more precise directions can be commanded.

But where can these vehicles operate in this transitional period?

At first driverless vehicles will be at level 3, with the 'I'm busy, I can't get to the controls at the moment' mode not yet an option. The driver will still be sat behind a steering wheel and actually using it in most cases. The first autonomy will likely be allowed on main roads in good conditions in the beginning with driver control on local streets. This could happen now as many makes of cars already have the technology fitted to do this but are restricted only by local laws. This may mean the learning of new motorway and A-road signs permitting such autonomy.

A mock up of a potential plate to be attached to the rear of a vehicle. In white with a red and white chevron border and the word Warning at the top. In the centre a large AV above the words Autonomous Vehicle. A graphic showing the gap between a rear and front of a vehicle shows brackets indicating a radio connection between the cars and a traffic light, interupted by an amber or green lamp. Below words explain Autonomous Operation When Lights Show and an indication that Amber means Prepared and Green means Operational
A suggestion for an illuminated plate to be attached to the rear of a vehicle to indicate to following vehicles that autonomous operations are in use

I foresee from this an interim period when autonomous vehicles have to display some sort of external evidence of potential auto control, probably backed up by an electronic 'black box' of trickery to meet certain criteria. Could there eventually be lanes designated only for autonomous cars, the outer lanes, geofenced to prevent access until you select autonomous mode? You might try to join in but the system just won’t let you. But when you can join in you could be cruising with cars travelling along at near to 200mph.

So does this mean that wise purchasers should be ticking these level 2/3 autonomous options on their vehicle builds now? Note to self when ordering that car, choose adaptive cruise control, lane departure and speed limit recognition camera option on the next build. It’s only another £4k after all. And when LIDAR becomes available I’m sure this will be just as, ahem, competitively priced.

This will all have to be developed in conjunction with the latest 5G mobile networking systems. This new high speed, high capacity internet will be needed to do the physical geofencing and authorisation along with the various car to car [v2v] and car to surroundings [v2b - b for base] communications needed for safe use of packed roads and high speeds.

In time it will be these same main roads where eventually all lanes introduce compulsory autonomous operation and older ill equipped vehicles will be barred.

And you can be certain that at all these stages governments and authorities will meet resistance from some, so expect a considerable amount of discussion and opinions. Mainly by me.

All this as we head towards full level 4 and 5 autonomy. But as I am considering the transitional period proceeding this then what else can we expect? And in particular what about the transition of our current cars.

At present the driverless mule cars being developed by manufacturers, the big tech giants and tech start ups resemble our current vehicles but splattered with an ugly array of cameras, radars, LIDAR, other sensors and devices all over them. In time this technology will be miniaturised and so seamlessly integrated into our current saloon, estate, SUV and lorry shapes.

It is also reasonable to assume that the developers of this technology will also want to sell it as aftermarket accessories to vehicles that don’t sport such stuff at present.

But this may not be the only offering because I predict the self driving robot. I discussed this in my comedy piece in February this year.

I foresee the introduction of approximately human sized self driving robots with the ability and technology to lock onto the necessary 5G systems and other relevant networks, connect to our cars using their inbuilt OBD socket then accurately survey and assess their surroundings. These robots would be able to actually clamber into our current cars and quickly communicate with the new tranche of driverless vehicles around them. Within they would have powerful computers and externally cameras, sensors and multiple ‘limbs’ to operate all the various functions of our current vehicles wherever the switches, seats, foot pedals, handbrakes, dials and controls are fitted. They may even be able to get out to change a wheel, check the tyres and even set the clock to British Summer Time as well. So clearly better than us.

I predict these robots will come before stage 6 autonomous cars are universally widespread, with licenses to roam wherever other autonomous vehicles may go.

Mind you there should always be a need to retain our own manual driving skills. I don't foresee most cars without any antiquated steering, speed or braking controls, even if they are usually tucked away out of sight. We'll need these to go where the maps don't map, such as into the wilderness, through temporary road diversions or into the depths of Morrisons' car parks.

And some specialists will still be needed to drive these auto-cars when things go wrong. I'm assuming the car will work out itself when to drive off to the garage for an oil service or get a flat tyre sorted but someone will have to pick up the ones that have faults reading something like error 404 bad sector. So don't put that Highway Code in the bin just yet.


Part Three - Issues and Scenarios

Within this third section I highlight just some of the issues surrounding this fascinating and complex subject.

One frightening aspect of autonomous control is the question of how a vehicle automatically relates to making life or death decisions. This is something we as humans already do when driving.

For example if you are driving along past some parked cars and a person, previously unseen, steps out from between two high sided vehicles in front of the car most of us would brake heavily to avoid contact. It’s natural. Usually before we can reassess whether our braking may affect any following driver.

Make that person stepping into the road your own child and you are likely to alter your decision, even swerving into the path of an oncoming vehicle with a chance of potentially killing yourself rather than the child. All without any thought to the opposing driver or occupants.

If asked to assess this in a calm and controlled manner, with enough time to work out all the permutations a different scenario may present itself. If you knew that hard braking would avoid your child’s death but knock them over without any major or long term repercussions, plus any following car was astute enough to brake behind you in time and the opposing vehicle had a number of other children on board who would all escape injury then your decision may change. To knock your little offspring over to teach them a lesson for being so cavalier in their attitude to road safety.

The trouble is we humans can’t work all this out quick enough. But we are building machines that can. So these machines have to have these sort of morals programmed into them.

There are many examples of these ethical decisions already out there in Internet Land so I won’t be repetitive here suffice to say they mostly base around speeding trains and pulling levers to decide which sub-branch line is selected thereby rendering different groupings of people being hit. The concluding moral usually being, choose to hit less people, select killing the elderly over the young and save people rather than cats. However, the true lesson to be learnt should be stop hanging around highly dangerous train sidings, particularly if you are an elderly singleton with your pet.

So let me propose some novel but more real world driving scenarios.

You are travelling in lane one at the legal speed limit on a multiple lane high speed road and arrive upon a joining junction. You notice two cars driving slowly up the slip road who will want to join your lane. They have not yet got up to the speed you are maintaining and you have plenty of room to pass by before they join the stream. You are also aware that no one is travelling along in the lane behind you but a fast moving [read speeding] vehicle is behind you coming up fast in the lane outside of you.

Suddenly, despite the adequately long slip road the first car joining the traffic makes a sudden and poorly executed swerve manoeuvre straight into your path right in front of you, seriously impeding your position and rendering you unable to brake in time. At the same time the speeding car in the outer lane is now blocking your escape route into any outer lane and the second car joining the traffic is blocking your escape route back onto the slip road.

You have no choice but to collide with one or another. Which car should you hit?

Take the opportunity to craft your thoughts as a comment. Take your time to consider all the implications. But if you are an autonomous vehicle you have 15 milliseconds.

Let’s try another. A little simpler this time and with no potential death.

An autonomous car is joining a motorway. A car, already on the motorway on the inside lane moves out to the centre lane to allow the autonomous car to join. They are now both on the motorway doing the legal top speed but travelling along next to each other. Should the autonomous car slow to allow the car already on the motorway to move back to lane one? And if it did and the other car did not move back to lane one what action should be taken by the autonomous one?

Again please feel free to comment.

A third scenario. Two autonomous cars, each equipped with a v2v system, approach a roundabout to arrive at the same time. Normal yielding rules apply on roundabouts but what if the one who has to yield has to wait a long time because the other has a long train of vehicles behind them? Logic may dictate via the v2v system to slow the approaching car with priority to allow all to progress the most efficiently. Or should the car with priority consider the stream of vehicles behind, who all technically have priority over the other.

Remember to consider all general occupants and the environmental impacts of your choices.

Now, add in a priority level. If one were an emergency vehicle or even if an emergency vehicle was in the train of following ones surely they would command full priority, no matter how much any other vehicles are impeded.

So, with priority level a consideration could the collective computers start addressing occupant needs? Is one occupant late for work? Is one just shopping [you can tell from my biased terminology that I am probably male. I am]? Do more occupants increase an individual vehicles' priority? Or even, has one driver had the benefit of more priority decisions this month?

And will autonomous vehicles make overtaking decisions based on all this? Deliberately slowing or stopping cars to allow others to proceed.

If all this is so, I think I’m registering my occupation as a Heart Surgeon and then filling my mobile phone’s calendar with fictitious operation appointments.

And so far I haven’t really touched on the subject of goods vehicles. These classes are the most likely to be fitted with this autonomous technology at first. After all the big lorries tend to travel major routes and often visit the same tightly controlled distribution depots.

So let’s set a scenario involving a number of autonomous heavy goods vehicles, all in an effective convoy, possibly cruising along, slipstreaming each other inches apart in order to travel the most economically.

What happens when you are cruising along in lane 2 in your non-autonomous car passing this effective wall of trucks on the inside lane but you need to exit a slip road ahead? What could you do? Should legislation determine convoys have to leave a gap between every, say, 5 lorries? Or should legislation dictate that 'convoys' completely break apart on approaching junctions to prevent last minute exiteers* attempting to fight for the same limited space? Or will we need a different road engineered solution, such as 'convoys' being restricted to outer lanes? So will we eventually have new junctions that pass over the carriageway and join from the other side?

As you can see there are a number of issues that are currently going through the minds of the programmers and engineers who are active in the field of autonomous vehicle operation and the more scenarios like this that we can imagine the better the autonomy will become.


Part Four - The Future

I have already addressed priority mode, particularly in relation to emergency priority, but what about our future day to day journeying?

I foresee an important setting to be made at offset is your own, settable, priority mode. Unless the car is singing the same tune as your calendar of appointments within your connected mobile device it will need to know the urgency of your journey. After all at times we are in no particular hurry and don't mind a leisurely drive. During other times economy may be our main driver, as it were. Maybe you wish to avoid tolls, or motorways. Then on some occasions you may be running a bit late and want to get on with things. Or there may be an emergency which you need to attend to and the vehicle would be instructed to travel as fast as is safe to do so.

In a sense we already have similar basic options on our current cars, or at least those sensible ones fitted with automatic gearboxes, although admittedly the actual speed is more dependent on the angle of your right foot. That is what the E-S-M [or similar] switch does in your car, it chooses your selected priority mode and adjusts the car engine and possibly suspension characteristics to suit. Typically, E for economy, S for Strewth this car can shift a bit and M for Memories wherin I remember when we had to actually choose the gears ourselves.

This of course means that when the autonomous vehicles are trundling around those cars set on the more leisurely settings should prioritise those on a speedier setting so every traveller is satisfied. Overtaking will occur in this future but only with the 'permission' of other vehicles. In fact you may notice that in some cases all cars going in opposite directions stop to allow another to overtake a whole queue. Even if you don’t spot this happening because you are resting in a catatonic state in your car, or perhaps reading the latest Rom-com, which amounts to much the same thing.

Also at junctions those cars set to economy or leisurely may wait for those with more urgent settings to pass by first.

This automated priority could be abused by some so don't be surprised if future legislation limits instances of hurrying to create a fairer system.

However, in time we shall become accustomed to this sort of behaviour without entertaining jealous thoughts of others.

But whatever mode we preselect all will be prioritised over the autonomous goods vehicles trundling around, looking for somewhere safe to drop their cargo.

Then there is the case of money. Isn't this always the case? Could wealthy individuals purchase priority? Maybe self appointed 'celebrities' will demand progress to avoid being caught up in a queue with the rest of us? And perhaps the most interesting question of all, what about the old geezer with a classic car?

By classic car I am thinking maybe a 2019 model. You know the sort. One that is not really autonomous at all. This ancient relic has no way of interacting with the then contemporary tranche of autonomous vehicles and will struggle to merge into a gap of fast travelling auto-vehicles seemingly joined nose to tail at high speed on the major roads. Well, fear not. For the rules of autonomy mean that autonomous vehicles have to do all they can to avoid accidents, so even a rogue, manually controlled one will have carte blanche to proceed as they wish and all the driverless ones will just jolly well have to get out of their way.

Now when the majority of vehicles are fully automated and controlled under a vast database of v2v and v2b [vehicle to base/infrastructure communication] systems, will we no longer need visible, plated speed control signs? After all the vehicles will know what speed to travel according to legislation and the road type. So, will top speed be effectively unlimited?

I think not. For a start there are consequences of potential accidents being more dangerous at higher speeds. Autonomy, whilst highly likely to reduce accidents, could not work to prevent them altogether. Mechanical failure, physics and build quality could all still play a part. Road traffic accidents can be minimised by risk management but no matter how much effort is put in they can still happen.

Finally, unlimited top speeds present questions from an ecological standpoint. Even if we have entered an age of unlimited free solar energy, because wear and tear on components would still apply.

So these are a few things we will probably have to look out for in a world of common autonomy amongst vehicles.


Part Five - Interesting Questions and Considerations

Let me ask a question that may seem silly at first but bear with me because it has a serious undertone. Will our vehicles eventually let us get into them in the first place?

After all we will programme them to protect us from ourselves. We will demand that these cars can take us to the pub and return us home when we ourselves are incapable of doing so without risk to ourselves or others. We will ask them to transport our nearest and dearest in the safest way possible. We will use them to transport our goods to destinations of our choice reliably and efficiently, without additional supervision. And the elderly and infirm will need to be able to fully trust these machines to protect them when they are unable to do this themselves.

To ensure the highest standards of safety we will programme them with the ability to self learn from errors made and their experience will be put to common use in vast databases to ensure the errors of one can be learnt by the many. In time this self learning will be more efficient within the databases than within human programmers so in essence the learning will supersede human ability. Some characterise this learning curve as becoming 'self aware'.

This could be a frightening issue for those that construct their mindset based on dramatic science fiction stories and who may foresee a future when these advance vehicles refuse to transport their owners because they judge their safety to be more important than the journey. And no journey can guarantee safety.

Or if we send them off to find a parking space will they definitely return to us when summoned back? Or will they consider the effort just too much to bother with? Or possibly will they be too busy picking up a more 'deserving' passenger?

This is all of course something that will not happen because we can, or at least I can, foresee this potential issue.

There is a fear amongst the pre-mentioned sci-fi followers to assume that a robotic future means that mankind will be made irrelevant upon machine self awareness. After all, some argue, if the machines are 'better' than us then why we would we needed? The answer to this is actually simple. Machines are made by mankind, for mankind and without mankind what is the purpose of said machines? I can work this out so I am sure robotic vehicle version 935.8.487 can figure this out as well. Even if it has to find and read this article first.

The above arguments raise another issue. Should we be able to trust future autonomous vehicles to transport our children and therefore at what age?

I believe this is no more complex than consulting current standards of childcare. In other words it is fine to send the car off to take your child to middle school but not send your new born baby fifty miles alone to be greeted by the grandparents.

The same would be for transporting your animals. By all means send Fluffy to the vet, provided the surgery is prepared to accept the consignment and Fluffy is secured in the vehicle with sufficient fresh air and water.

And on the subject of transporting goods this will become commonplace, with the vehicles secured at offset and only accessible by the appropriate person on arrival.

All of which is ideal for drug running businesses across the border. The authorities never suspecting this, mainly as they will rightly assume all the drug transport will be via autonomous drones. But what if the drug vehicle carriers are impounded? Who is deemed legally responsible, the vehicle owner but it could have been stolen, the manufacturer or the software engineer?

Sorry, I have swayed into story time again. Anyway it's fun so let's continue with some other radical thoughts.

What about a future where autonomous cars, who’s owners have died or abandoned their vehicles, still roam the streets? After all that is their raison d'être. Forever left to search for electric charging points, heading off to get serviced and driving around with no passengers aimlessly searching for a reason to exist? If this isn't a side feature in the next Ridley Scott movie I will be extremely disappointed.

And finally, could self driving cars communicating together end up doing formation patterns, just for the sheer fun of it all? I hope the programmers include this possibility. Look out for future photographs of busy motorways from above wherein the cars are precisely positioned to spell out 'vinceunlimited'.

And that's neatly back to where we started off. So, while you still can, drive safely** Vince

I’m sure I will return to this fascinating and developing subject in the future so keep following me on this web site, on WordPress, Twitter and in an appropriate lane on the A31 for more insightful commentary.

Related Links

Author: Vince Poynter
Version 5.267 28 Jun 2019
Published over five days during week commencing 24 Jun 2019
My related blog post, According To Me, can be found by scrolling down
The whole blog can be found at http://vinceunlimited.co.uk/blog.htm or by clicking the blue button link above
My related podcast can be found by searching for 'Pod 005 - Lanserguided' at http://vinceunlimited.co.uk/podcasts.htm or by clicking the blue button link above
My short stand up routine about driverless cars was performed on 20 Feb 2019 and can be found on YouTube at https://www.youtube.com/watch?v=7g5UuzeIm-M or by clicking the blue button link above
Wikipedia Autonomous Driving Information Source: https://en.wikipedia.org/wiki/Autonomous_car\
The photograph is of a Benz Patent-Motorwagen and was taken by the author at the Beaulieu Motor Museum, Hampshire on 17 Aug 2019
The Autonomous Vehicle Plate was designed and drawn by the author. It was created by adding the text into a Keynote slide, then printing and hand drawing the lines and graphics
* Note: This is not a Brexit reference
** Message also addressed to all future autonomous vehicles



June 2006

According To Me

My first mention of autonomous driving and the insurance implications...

A close up photograph of toy green classic mini in a rather tatty state with red overpainted opening doors, bonnet and roof with a blue circle logo on the door
The result of an accident between a car and a small child

I have just read about a development of a technology from one major car manufacturer that encompasses radar, cruise control and the ability to follow white line markings whilst steering to effectively allow the car to drive itself.

All these technologies are already produced but this car combines them all.

The car in question is a Honda Accord - the pensioners of Britain must be wetting themselves with glee.

All this relies on effective road marking of course but nobody has yet made that quantum leap into the future to envisage who might have to take responsibility should it all go pear-shaped.

Can we look forward to the accident case where the driver claims that he was not actually controlling the car, whereas the manufacturer will be pointing to some small print in their instructions whilst the insurance company attempts to blame the road maintenance companies?

All of which means the poor motorist that was crashed into will be a pensioner himself before he gets compensation.

All of which he'll spend on a new Accord.

And the circle will continue ad infinitum…

Author: Vince Poynter
Added to the Autonomous Driving Page in version 5.267 28 Jun 2019
Version 5.125 1 Jun 2018
First Published: Blog within Version 2.03 14 Jun 2006