Autonomous Vehicles: From Science Fiction to Road Reality

My first encounter with autonomous vehicles (AV) was through the 80s TV drama series titled "Knight Rider" starring David Hasselhoff who holds the Guinness World Record for being the most watched man on TV.

No, David Hasselhoff did not invent the Autonomous Vehicle. However, like what Michael Jordan did for NBA, David was back then the poster boy for AV and for the last 3 decades, technology has finally caught up with science fiction. AVs are no longer a fantasy but more and more, we will be seeing their rollout.

Through this article, I want to take you through the journey of autonomous vehicle development and end with my prediction for the future.

History

You probably think AV came about in the 70s since Knight Rider premiered in the 80s. Truth is, AV was a concept that was given birth in 1925 by Francis P Houdina, an electrical engineer in the US Army. Using radio equipment, he controlled and operated a car through the streets of New York City. That was the start of AV and as technological breakthroughs came about, so did the viability and ambition behind AV.

The Development and Benefits of AV

Over the years, the development of AV has garnered momentum and in 2009, Google started a project known as WAYMO to look into self driving cars which according to the latest update, WAYMO self driving cars will be rolling out on the streets by end of 2018.

How autonomous will they be? In AV, there are different levels of autonomous-ness. (Read: I think the correct term is automation)

Source: SAE International – Levels of Driving Automation

As you can see, Level 5 is the ultimate goal and it means there is no need for any manual intervention at all. The passenger can take their eyes off the road and be engaged in whatever they want to do. It is a totally hands-free situation and one I know will take more time to achieve.

Despite Level 5 being the 'dream mode' which companies are still falling short but driving for, autonomous driving at any level of automation presents a less tiring daily commute that helps drivers conserve energy and avoid accidents due to human errors. Statistics show 94% of driving casualties resulted from human errors! The remaining 6% comes from road and weather conditions. Given this stark statistic, the decision to release the steering wheel and brakes to a robust automation system may not seem that risky after all.

Yes, all this is on the assumption that the system is well designed, backtested to the core and certified robust for use. We know no system is infallible and there will be hiccups and glitches. However, in view of its potential benefits, some of which are already being enjoyed by the masses such as assisted parallel parking, it is definitely a direction worth pursuing by car makers and tech companies.

The Challenges

Every technology has its flip-side. Self driving cars are no exception.

#1 Right from the bat, getting the AV to work perfectly (Level 5) on its own is HARD. Even the CEO of WAYMO put up such a statement recently which I don't think was meant to discourage the industry but rather a reminder of how challenging it is to move from say Level 3 to Level 5 (i.e Full Automation)

Waymo CEO: Autonomous cars won't ever be able to drive in all conditions. And it will be decades before self-driving cars are all over the roads, John Krafcik says.

#2 Regulations are going to be a hardcore issue down the road. Definitely, there will be laws and regulations to govern the deployment and use of fully autonomous vehicles. As of now, most AV are not 'driverless'. The driver can switch off and turn their attention to something else...for a while but they still retain control over most of the responsibilities of guiding the car to its destination.

Things will change when full fledged AVs hit the road and this is not just going to affect the transport authority in each country but also related industry. For example, I can imagine the insurers going crazy and trying to debate with each other which 'driverless' car was in the wrong! Or maybe this won't happen since they are supposed to be fully automated and free of human errors.

Anyway I digress but this is just such an interesting scenario.

#3 The Search for AI Professionals is challenging because...

According to the study, compiled by the Tencent Research Institute, there are just 300,000 “AI researchers and practitioners” worldwide, but the “market demand” is for millions of roles. Dec 5, 2017

The development for AV has several core engineering requirements, among which AI or Artificial Intelligence plays a major role. AI is aptly the brain of autonomous vehicle and is supported by other core engineering disciplines such as machine learning, vision engineering, big data and cyber security.

To build the future for AV, the investment into AI work is of paramount importance and anyone who is an AI professional is in HOT demand and the same goes for the other core disciplines mentioned above.

Future of Autonomous Vehicles

Will autonomous vehicles take off or will they be grounded to a halt as a result of the challenges above and of those not even thought of?

As vehicles become increasingly connected, greater applications and benefits will come about. Needless to say, problems will tag along. One of the key concerns for fully connected vehicles (IoV) is that it will open up the gateway for car hacking and that is a very serious issue not to be taken lightly.

When a computer or phone is hacked, personal data and privacy are compromised and yeah, that's bad. However, when your car is hacked, it can be a matter of life or death.

In the face of such cyber security concerns, the only way is not to downgrade and disconnect. Instead, there is a demand for cyber security specialists to make this work out and to ensure the ride is not just enjoyable but absolutely safe as well.

#AutonomousVehicles #AI #MachineLearning #DeepLearning #CyberSecurity

Published by: JK Chan

Practice Lead / Partner, TAP