The future of autonomous driving in company fleets and the mobility budget

Thanks to an initiative organised by Espaces-Mobilités, and under the guidance of Xavier Tackoen, I had the opportunity and honour to see and test autonomous vehicles operating in China in three cities of approximately 20 million inhabitants each: Beijing/Peking, Shenzhen and Guangzhou/Canton.

In all three cities, these vehicles transport passengers within urban areas defined by the local governments, which supervise each vehicle in real time.

Of course, these vehicles are also supervised by their respective operators to ensure smooth traffic flow in the event of an unusual problem. An intervention may, for example, be necessary to authorise the autonomous vehicle to perform a manoeuvre that would normally be prohibited, such as crossing a solid white line when a lane is obstructed.

We are talking here about a few hundred or a few thousand vehicles per operator. For passengers, the conditions are very real: they book their journeys via an App and pay for their trips. For the operators, however, these are still tests conducted under real-world conditions.

For mass deployment, millions of robotaxis would be required. However, the Chinese government is not necessarily seeking such a rapid development: it pursues a policy of full employment and has enough human drivers available.

First of all, it is important to clearly distinguish between assisted driving and autonomous driving

With assisted driving, the driver remains legally responsible and is the policyholder: this is referred to as Level 2.

This is, for example, the case with Tesla vehicles equipped with their “FSD”, which have recently been authorised in certain European countries, including Belgium. Here, there is no question of reading the newspaper behind the wheel.

For autonomous vehicles, the company that puts the vehicle into service is legally responsible and is the policyholder: this is referred to as Level 4 or 5. Unlike Level 4, Level 5 no longer has a steering wheel or pedals.

Note that Level 3 is a mixed model in terms of responsibilities. It has not really been successful and is being abandoned by almost all players.

In terms of autonomous driving, China is several years ahead of Europe and, in some respects, ahead of the United States. But there are many different strategies.

For example, there is the champion of vertical and global integration, Tesla, which masters all areas of expertise, versus a partnership strategy in which each party contributes its own expertise and added value and integrates it into a vehicle.

The areas of expertise are numerous: vehicle manufacturing, components, telecommunications, processors, cloud, artificial intelligence, certifications, maintenance and operational structures, commercial networks...

Europe is still developing relatively little of its own expertise in autonomous driving. For the time being, it has notably adopted a strategy of simply purchasing vehicles produced by China or the United States. At the same time, however, European manufacturers are gradually integrating the technologies that will make their vehicles autonomous.

European car manufacturers are therefore choosing their partners both in the East and in the West.

China and the United States are consequently engaged in fierce competition, side by side, with enormous amounts of capital committed.

The main players

For robotaxis — in other words, driverless cars — the leading players in the United States include Tesla and Waymo, while in China we find WeRide, Huawei, Didi, Baidu and Xpeng. But there are others.

Today, most of these cars are considered Level 4, because they still have a steering wheel and pedals. Within a few years, they will evolve towards Level 5 vehicles, without a steering wheel or pedals.

They are fully autonomous, from one parking space to another, even outside public roads.

Both countries are currently in a phase of intensive testing, with several thousand vehicles operating in areas authorised by local public authorities.

We are very close to the mass deployment of large robotaxi fleets.

It is nevertheless difficult to award first place in this competition, because we are still only at the beginning. It will be more of a marathon than a sprint: there are still many competitions to win before we can crown the world champion.

More and more vehicles today are equipped with cameras and sensors that assist driving.

These vehicles are also connected and allow manufacturers to prepare for autonomous driving: they read the roads, build maps and create a library that feeds the operators' artificial intelligence engines.

Some Chinese production vehicles are already equipped with so many sensors that they would require almost no modification to become autonomous.

We are not far from activating Level 4 through a simple software upgrade and a subscription to an operator's supervision service.

Breathtaking prices...

Let's take an example: a Chinese SUV comparable to a Volvo EX60 costs between €15,000 and €25,000 when purchased in China.

To compare this with a European price, however, the amount needs to be adjusted to compensate for the tax advantages these vehicles receive from the Chinese government, which represent between 15 and 35%.

This brings the price of the same SUV to approximately €33,000.

Fortunately for our European manufacturers, these vehicles are not — yet — imported into or manufactured in Europe.

And what about autonomous buses?

In Belgium, as far as public transport using autonomous shuttles or buses is concerned, an autonomous shuttle operated by De Lijn is in operation in Leuven. It is a European first, led by Tim Asperges together with Espaces-Mobilités.

This shuttle operates on a route right through the city centre that is particularly complex because of pedestrians and the very large number of cyclists. In Leuven, the bicycle is king.

But all of this is still a very long way from mass deployment.

Many factors still need to come together. Here are some of them:

• An enormous and rapid vehicle production capacity, requiring capital on a scale that is no longer that of a simple “scale-up”.

• A range of vehicles that genuinely meets users' needs: price, brand image, trust, usage...

• Cost-effective production of sensors, processors, software and artificial intelligence technologies.

• Acceptance of the technology by users. I am not referring here to insurance, because these vehicles will be readily accepted: they have a very low accident rate and even fewer at-fault accidents involving injuries. They are much more reliable than humans, with statistics to back this up.

• Acceptance of the technology by local authorities, which will combine licences, quotas and the definition of the areas in which they will allow these vehicles.

• A shortage of human drivers — an obvious factor that will facilitate their acceptance.

• A well-established sales, maintenance and supervision network, both locally and globally.

• Millions of kilometres driven in all weather conditions over years of testing in order to prove the reliability of the vehicles.

• Local operational managers.

In short, you will probably reach the same conclusion as I have: their widespread deployment across all Belgian cities and suburban areas will not happen this decade.

But the robotaxi has its Trojan horse

It is Level 2++ assisted driving, which is already on our roads — Tesla FSD, for example — where the driver barely touches the steering wheel or pedals but remains legally responsible for driving the vehicle.

The first people travelling in these autonomous cars are private individuals who have the means and the motivation to buy them.

Soon, if it has not already happened, we will see autonomous vehicles appearing in company fleets, provided that the price of this “option” remains within employees' budgets.

What role will Level 4 or 5 autonomous vehicles play in company fleets?

By definition, an autonomous vehicle (taxi or bus) is shared and operated by a company external to the businesses using it.

These vehicles should therefore be seen as an alternative to company cars currently allocated to specific employees.

They will therefore be shared cars available through Pillar 2 of the mobility budget.

But we will still have to wait many years before this solution is genuinely available on a large scale.

To be continued, but there's no rush ;-)