The Future of Mobility is Autonomous

Last week I was fortunate enough to spend three days driving a leading German manufacturers “autonomous car” for 1,500km (950 miles) across California and Oregon. The experience was far from perfect, but I was massively impressed with a concept that will no doubt eventually go mainstream, and disrupt the motor industry. 

As I’ve previously noted: the mobility industry is being fundamentally disrupted/improved/changed forever by three innovative tech-fuelled pillars:

1.    Autonomous vehicles

2.    Electric vehicles

3.    Ride hailing/sharing

Thanks to these pillars, we will experience the greatest shift in mobility since a guy by the name of Henry Ford replaced the horse with a Model T. 

I’ll speak only of autonomous cars here... 

It is worth noting that there are several levels of “autonomous cars” with many convoluted descriptions from the marketing teams of auto manufacturers. (much like the 5G buzzword in telecoms). The car that I drove was far from being a Waymo (Google) fully autonomous vehicle – indeed, it was not even the consumer-market-leading Tesla Model S/X, but it fully adhered to Californian law by complaining bitterly if the driver’s hands were off the steering wheel for more than 30 seconds or so. That said, whether out on the open roads or in rush hour traffic, it literally drove itself...

What I liked about the experience:

* The seamless handover (and/or offered steering assistance) between the computer and driver.

* The smooth transition between computer and driver during accelerating and braking events.

* The ability of the car to emergency brake and swerve in extreme cases. This feature alone could save many lives, particularly with fatigued drivers. Due to this support, I found myself far less tired after a long drive as the computer did a lot of “thinking” on my behalf. 

At times I was not sure who was in control, myself or the computer… but when the car emergency brakes in traffic before I even had a chance to react, I knew who was really the boss.  

* The way in which the car can be “paced” to a leading car, to the exact inch in distance and irrespective of the leading cars acceleration / breaking behaviours. I can imagine this was somewhat creepy to one particular driver whom I followed for 50 miles through twisting Oregon roads without any variables changing. With a human’s irrational behaviour being the number one reason for congested rush-hour traffic, this could help massively in reducing city rush-hour drive times - add “smart cities” of the future to the mix, along with their connected traffic control systems, and one can only imagine how this will be a game-changer for mobility.

What I didn’t like about the experience:

* As the major navigation/input readings of the vehicle are road markings, it has a tendency to “panic” in potentially dangerous situations such as conflicting lines on the road in construction areas, as well as wet roads and bridges with no markings.

* The cars systems are a tad “overeager” on tricky corners where the preset speed is high and thus maintained... it clearly backs itself at least.

* The car had the ability to back off and get itself into a lull when traffic is heavy. More advanced autonomous levels - including Tesla’s - have better “lane assist” functionality.

Where to from here?

Although the vehicle that I was driving was guided by a lower level of automation - one could call it semi-autonomous, due to the driver still having to take control - the potential is clearly there to see. 

Key going forward will be increased AI- capabilities, with machine learning (ML) in particular being a game changer.  

With this in mind, I believe that traditional investors are underestimating companies such as Tesla (as well as Google’s Waymo) even though they are clearly leading in tech, as well as autonomous ML. Consider this… every Tesla now has at least six cameras on it, all recording real world simulations which can be “uploaded” back to Tesla at night while you sleep. Tesla then ingests the data (scenarios) into their machine learning algorithms, where the AI can basically teach itself how best to react to every situation - such as if there is a child and a dog in the road at the same time...

Hold on tight folks… the next decade is going to rock personal mobility as we know it.