There are already plenty of devices without a steering wheel! Think about the old Roomba, the autonomous vacuum cleaner introduced back in 2002 by iRobot. Well, it bumped on chairs and walls but that was not a big deal. Upon bumping it backtracked and changed its direction: no problem! Subsequent releases learnt to recognise obstacles (including cats, although I am pretty sure cats were able to spot an approaching Roomba well before Roomba spot them). I mention Roomba because it was probably the first mass market fully autonomous vehicles and it embedded an amazing (for the time) set of sensors and smart (intelligent) software.
Of course, roaming in a flat with licence to bump here and there, is quite different than roaming in a city, at much higher speed and with no bumping permission. Hence, no surprise that fully autonomous vehicles are still scarce. Scarce but real!
Zoom, by Amazon, is an example of an autonomous vehicle designed from scratch. It has no steering wheel (like the prototype GM car in the picture) and I found interesting hearing the impression of its passengers (watch the clip). I already tried a number of self driving cars (in dedicated spaces, not open to normal traffic) but most models I tried had a steering wheel and even if I did not touched it during the (short) drive I found that its mere existence gave me a better sense of security. The only one without a steering wheel was the one in Masdar, Abu Dhabi (it looked a lot like Zoom). It didn’t give me the creep mostly because it was driving underground in what looked like designated lanes. I am pretty sure I would feel uncomfortable experiencing a normal car without a steering wheel in normal traffic, like I am missing something (important). My feeling is that we will have to wait a few years before fully autonomous cars become common, and even more before they lose the steering wheel. A different story applies to VTOL -Vertical Take Off and Landing- taxi drones that should be starting service in some areas (Dubai, Singapore) already this year at a fee, this is the interesting part, that compares to the one a a normal taxi during peak hours.
On the contrary, fully autonomous vehicles for good transportation, both land and air transport, are likely to become common during this decade to the point that we won’t be paying any more attention to an autonomous delivery of pizza by the end of this decade.
Based on the FTI’s report we can expect, also fostered by the pandemic lockdown in several areas, an uptake of last mile delivery based on autonomous vehicles. JD.com, a big Chinese eCommerce, has delivered over 13,000 packages covering close to 7,000 km in lockdown areas during the pandemic. Clearly they had benefited from much lower traffic but still it has been quite a trial and learning experience. According to the WEF, the surge in eCommerce will lead to a 36% increase in delivery vehicles by 2030 with last mile delivery increasing by 78%. This will increase pressure in developing effective autonomous vehicles for goods transportation.
The increasing use of autonomous drones for infrastructure inspection will stimulate evolution in image recognition and related AI. In turns this will result in better “follow-me” capabilities (already available in mass market drones for photography). It is not all bright. As we are seeing on recent newspapers articles, and pointed out in the FTI’s report, privacy concerns are rising against police using autonomous drones and robots for patrolling and following people. The recent EU document on ethical use of AI is explicitly classifying the use of surveillance based on AI on autonomous vehicles as a High Risk application.