Automated and Electric Vehicles Bill

Written evidence submitted by Anthony Beken (AEVB 01)

In the headlong rush to a brave new (driverless) world, it must be born in mind that there are many problems that will need to be addressed:

1. Human Interaction
Autonomous cars struggle to recognise humans alongside the vehicle or walking in front of it and have a limited field of vision. Pedestrians, cyclists and workers in the road could be put at risk of collision. For example, an autonomous car is travelling along a road. A police officer standing by the kerb is signalling for traffic to stop. The car fails to recognise him, carries on driving and causes an accident. Form points out that police officers often wave cars through red lights, or stop them at green lights. How would a driverless car cope with that situation?

2. The Weather
Poor weather leaves most automated vision-systems in the dark and unable to see. Snow, sleet, ice & driving rain are untested hazards for self-driving cars, and vehicle performance is uncertain. It’s unknown how autonomous cars will handle, avoiding ice, detecting lanes under snow & driving in wet conditions. The car might well detect snow as an obstruction and refuse to move at all.

3. Morality & Ethics
The law and ethics can conflict with one another in dangerous situations – the right choice might require illegal acts. Legalist safety directives could lead autonomous cars to make poor decisions in emergencies. For example, it is unknown how autonomous cars will handle choosing between an emergency stop; which would cause a fatal collision to those behind; or carrying on and running down a child in front. What choice would the car make? Software may spot obstacles ahead, but finds it hard to distinguish dangerous ones, or even which ones are socially acceptable to ignore. Form suggests that human drivers would not stop to avoid hitting a rabbit – but they might for a dog.

4. The Law
Safety testing must be completed outside the R&D environment in a statistically significant way. Legal theorists will need to account for the new problems raised by autonomous cars. Suppose the car detects the child a child suddenly appearing in its path, the car performs an emergency stop, the child is still hit but the car had performed exactly as it was programmed to. Who will be held accountable?

5. Security & Driver Safety
Technical errors on the road could have fatal consequences for passengers in self-driving cars. Legal theorists for autonomous cars will need reviewing to ensure that the cars make the best possible judgements. Metallic chocolate bar wrappers have caused driverless cars to panic. For example, the car detects an obstruction ahead and a sudden stop will cause an accident. The car has a directive to avoid collisions with objects. It also has a directive to obey the law by keeping lanes. Confused the car decides to sop – an accident ensues.

6. Cost & Affordability
Prohibitive costs could prevent autonomous cars becoming a common fixture on the roads. Competitive pricing will be needed if manufacturers are to gain a market foothold. £4,500 – £7,000 will be added to a car’s price tag for self-driving technology in 2025 (IHS Automotive). This figure will drop to £3,000 in 2030 and £2,000 in 2035 – but we all know how unlikely this may be.

7. Mapping
Every street, pavement and driveway that will be used by a self-driving vehicle will need to be mapped. To give a sense of scale of the project, there are an estimated 250,000 roads in the UK alone & new roads are being built and re-sited on a frequent basis.

8. Hearing
Even the most advanced prototypes are having a hard time interpreting emergency sirens. If they are ever to be allowed onto real streets, autonomous cars will need to be able to respond appropriately if an emergency vehicle tries to pass.

9. Reversing
Most self-driving cars cannot yet reverse without human assistance, although this shouldn’t be a problem to master, it’s not a priority for developers.

10. Insurance
Quite how insurance will work still remains unclear. When a driverless vehicle crashes, who takes the blame and who coughs up? This is still being debated and is a major barrier to letting them loose on our highways.

The answer proposed in the Bill is that insurance companies should – unless you've modifying the car's software or not installed "safety-critical software updates". The intention behind the Bill is to emphasise that if there is an accident the compensation route for the individual remains within the motor insurance settlement framework, rather than through a product liability framework against a manufacturer. Manufacturers will more or less be exempt from routine liability claims because of the bill's provisions allocating blame to either drivers or insurance companies. By shifting the liability for fault from manufacturer to insurer sends the wrong message to those manufacturers with less than scrupulous quality control procedures.
What if there is an accident when your car is driving itself? The comprehensive insurer would be obliged to compensate the innocent third-party victim as well as the policyholder. It would then be up to the individual insured to recover the value of their lost no-claims bonus from the product manufacturer, an almost impossible task for the average motorist.

11. Public Perception
If recent surveys are anything to go by, developers have failed at the first hurdle as it seems much of the public (myself included) is actually too scared to even step foot in one of the cars. Half of Britons responded that they would be unwilling passengers in driverless cars over safety concerns, with 16 per cent professing to feel 'horrified' by the notion of unmanned vehicles driving on British roads. Four in ten would not trust an autonomous car to drive safely, believing it would jeopardise the welfare of drivers, cyclists and pedestrians, a survey from price comparison site found.

12. Handing Over to a Human Driver
Pioneering driverless systems work on the principal that a human driver is present and will take over when required. But cars might struggle to give the human driver enough notice if & when they were needed. If humans need to take over when the software can no longer cope, they will need to be skilled drivers – as well as being on standby (not asleep, or over the limit for instance). This negates some of the usefulness of self-driving cars.

13. Hacking & Computer Error
Sadly, anyone who has ever owned a computer knows that they are not perfect. They can and do ‘crash’ usually at the most inopportune times. Manufacturers have been rushing to turn modern cars into smartphones, offering more & more remote options. There have already been several documented successful attempts to hack into these automotive systems and take over key functions of the vehicle. It is predictable that driverless vehicles will present too tempting a target to resist for the would-be terrorist and/or hacker bent on causing mayhem. There have already been a number of 'near-miss' events involving commercial aircraft and amateur operated drones and it is only a matter of time before a collision ensues, which no amount of airport security would prevent.

A terrorist on a motorway bridge with a ‘device’ could potentially cause major loss of life if they were able (& they surely will one day soon) to override a cars own control systems.

Also included in the bill are future provisions to allow government to force petrol station operators to install electric and even hydrogen car charging points. Current (no pun intended) electric vehicles can take up to 22 hours to recharge so the almost forgotten forecourt queues of Winter 2000 will reappear in spades if this requirement is ever enacted into law.

At the recent COP21 climate change conference in Paris, an agreement was reached to wipe internal combustion engines (ICE) from the face of the planet. Or at least part of the planet. The agreement stated that, by 2050, all new cars sold in eight US states, and five countries including the United Kingdom must make zero tailpipe emissions - which means no more internal combustion engines. This is not the vehicle (sic) to argue against such foolishness, save to say that I for one, will not be venturing north in the knowledge that I will have to stop, at least once, overnight, in order to reach Birmingham with existing battery technology. Makes HS2 almost palatable.

Today’s electric models are averaging 185 miles between charges and charging times are up to 22hours with a Level 1 charger and 4-6 with a Level 2 charger. There are no indications that technology will dramatically increase battery capacity for the foreseeable future, although DC Quick-Charge (DCQC) or Level 3 Charging is the fastest charging option available for EVs today. This is not a charging plug that you can install in your house because they cost up to £100,000 and need more power than the whole house. However, these charging stations offer about 40 miles of range in just 10 minutes of charging. Best for on-the-go traveling, DC Quick-Charge charging stations use 480 V of voltage and a typical saloon will take only 30 minutes to go from 0 to 80 percent charge. Level 3 charging stations are designed to charge the car only up to 80 percent because EVs can lose capacity from this type of rigorous charging. However, this is the best quick recharge option. Difficult to require your average petrol station owner to make the £500.000+ investment and even harder to imaging how our hard-pressed national grid will cope with the Government’s stated goal of 100% electric vehicles by 2050.
Perhaps we will all have to use a calor gas burner to make our cuppa after the news at 10.

Electric vehicles are zero-emission at point of use. However, emissions are produced during the generation of electricity, the amount depending on the method of generation, and the recycling of used batteries which are extremely ‘dirty’ in green terms. Therefore, the emissions need to be considered on a lifecycle basis so as to include power station emissions. In 2014, total electricity production stood at 335 TWh (down from a peak of 385 TWh in 2005), & was generated from the following sources:

Gas: 30.2% (0.05% in 1990)    Coal: 29.1% (67% in 1990)
Nuclear: 19.0% (19% in 1990)    Wind: 9.4% (0% in 1990)
Bio-Energy: 6.8% (0% in 1990)    Hydroelectric: 1.8% (2.6% in 1990)
Solar: 1.2% (0% in 1990)    Oil and other: 2.5% (12% in 1990)

So, at present only about 12% of our electricity generation source can be said to be green.

Instead of (or as well as) sounding the death knell for the internal combustion engine, it would behove the Government to act meaningfully to ensure the current emission standards & limits are actually being met by car manufacturers (& others) who market their products here.

October 2017


Prepared 1st November 2017