In Knight Rider, the American TV hit of the 1980s, crime fighter Michael Knight was the main man. He wasn't the star of the show though. That honour went to an inanimate object, KITT – a virtually indestructible and artificially intelligent, black supercar with high-tech dashboards, X-ray surveillance systems, and a sexy red strip light that flashed back and forth across its grill.
Despite his dashing good looks, actor David Hasselhoff didn't stand a chance as KITT invariably drove itself to the rescue whenever his character Knight got into trouble.
Less than four decades on and it seems that some of the concepts dreamed up by Knight Rider creator Glen Larson, who died in 2014, could soon become reality. It's a shame he didn't live to see it.
Autonomous vehicle technology is touted as one of the biggest investment opportunities of the 21st century. Still considered largely science fiction just a decade ago, the revolutionary concept started to take shape in 2009 when Google began beta-testing the world’s first driverless car, Waymo.
But it's only more recently, as major car manufacturers like Tesla, Nissan and Ford have joined the race and unveiled roadmaps to develop fully autonomous vehicles, that the excitement around autonomous vehicles has really taken off around the world.
The concept of an autonomous car is not something new. Most modern cars are equipped with features such as parking assistance, anti-lock braking (ABS), and warning systems that are automatically triggered and do not require any human control.
Car makers like Tesla and Volvo have designed semi-automated cars equipped with advanced navigation systems that allow automated cruise control and steering in certain road conditions, and which help drivers to navigate highways and parking lots. Still, a driver is required to monitor the vehicle at all times.
But now automakers are in the race to develop cars that are fully automated, involving no human intervention at all. Based on the scale drawn up by the international Society of Automotive Engineers, that is a vehicle with Level 5 automation (see diagram) that can drive itself anytime, anywhere, and under any conditions.
TRANSFORMING THE VALUE CHAIN
A common misconception about the autonomous driving revolution is that it affects only the traditional automotive value chain comprising mainly automakers and their suppliers. But the impact is likely to be far more wide ranging, industry experts believe.
“The future auto-driving ecosystem will involve cross-sector collaborations between the government, automakers, technology, and telecommunications companies,” Yong Huan, executive director of Intelligent Vehicle Technology Center at Chery Automobile, told FinanceAsia.
Chery Automobile is a Chinese, government-backed manufacturer that is in the process of developing the country’s first autonomous car in partnership with technology giant Baidu.
Mechanics followed by electronics have been at the heart of the automobile industry. But, as with other industries, they are likely to have to make room for the digital world as the sector enters the driverless era.
That is because self-driving cars have to be supported by next-generation technologies like big data analytics, artificial intelligence, and cloud computing. Self-driving applications will need to analyse a huge amount of technical data just to “see” the world immediately around it.
Automobile experts estimate that a fully autonomous car will generate 4,000 gigabytes of data each day – equivalent to about 1,000 high-definition movies – and that it will take more than an hour to transfer all the data even at the fastest speed under a 4G network.
And then there is the process of organising and analysing the data and then turning it into actionable insight, which would allow a vehicle to act without human intervention. This will involve machine-learning algorithms that enable a driverless car to forecast possible changes in its surroundings, or cloud-computing technologies so it can calculate how fast other cars are approaching.
“A lot of chip manufacturers are now taking the centre stage of the automobile industry,” Chery’s Huan said. “Their technology plays a dominant role in developing new cars and many of them are increasingly the price-setters for newly made cars.”
As such, the autonomous driving revolution is open to companies across various industries, which in turn will help to accelerate the sector's development.
“The [autonomous car] industry is progressing much quicker than we had expected,” said Yuanxing Lv, a director at Nio Capital, the private equity fund behind Chinese electric car maker Nio. “This is because the industry now embraces a lot more participants, like manufacturers of sensors, computing devices, chips, and more.”
RESHAPING THE WORLD
The wider human impact of the coming driverless car revolution is unprecedented and could potentially reshape entire economies by transforming the way a society operates and the way people work and live.
The most direct impact will be on the healthcare industry. In the future, hospitals and clinics will barely need to earmark resources for patients with car accidents because driverless cars are a lot safer – they will never get drunk, lose concentration or exceed speed limits.
“Autonomous driving is all about safety,” Scott McCormick, president of the Connected Trade Vehicle Association, said. “In the US we have 32,000 deaths from car accidents every year. In China, a person dies in a car accident every four minutes. With autonomous driving technology, these numbers can be massively reduced.”
Since 2009, Google’s driverless car Waymo has travelled more than 6.5 million kilometres, which would have taken an average driver 300 years to complete. It has so far caused 26 accidents and no injuries have been reported.
However, concerns over traffic safety heightened after Uber's driverless cars reported the first fatal accident in late March, causing the company to suspend its road tests across the US.
Driverless cars are set to lead to sweeping changes across the global jobs market too. Taxi and bus drivers will no longer exist once fully autonomous cars are in place. Pizza takeaways will no longer be delivered by a delivery man or woman and driving instructors will no longer be needed.
Commuters will also have more spare time as driving becomes more and more optional. Analysts at investment bank Morgan Stanley have estimated that driverless cars could yield up to 400 billion hours of extra free time per year, which could translate into enormous business opportunities for e-commerce, music and games, as well as other forms of entertainment that can be enjoyed while travelling.
Some industries such as auto insurance will shrink and even perish. Parking sites will be greatly reduced because fewer cars will be left idle. New industries like automobile analysis and software programming will emerge.
NOT A SMOOTH RIDE
There increasingly seems to be little doubt that the growing driverless technology bandwagon will eventually reach its destination. But before that there are a few bumps and roadblocks to negotiate.
For one, there is a lack of a unified system to govern the way artificial intelligence behaves.
“Imagine four cars running on the road and they meet. There needs to be a set of rules for them so that when one of them turns left, the other three know how to react in order to avoid collision,” McCormick said. “In order to achieve this, there is a lot of research and cooperation needed between different parties.”
Facing the unknown in many respects, car manufacturers and artificial intelligence companies are developing systems in different ways, which adds to the difficulty of putting together a master set of rules.
“It is a big question when it comes to data sharing because no one is willing to take the lead,” said Ping Zhou, vice president of Banma Technologies, a joint venture between Alibaba and SAIC Motor that focuses on developing internet-connected cars.
Another potential challenge is to develop sufficient computing and calculation capabilities for artificial intelligence (AI) to organise, process, analyse, and understand the massive data it collects every second.
Imagine it is snowing heavily and there is traffic jam ahead due to a car crash. "In this case, an autonomous car will have to analyse not only hardware data generated from all the sensors, but also real-time information from traffic departments, weather forecasts, public transport as well as satellite and mapping information to determine how to reach its destination,” Zhou told FinanceAsia. “It is another big challenge teaching AI how to analyse data from various sources.”
The ever-changing road and weather conditions mean AI needs to possess deep-learning capabilities in order to react instantly. If so, we're still some way from that.
“AlphaGo [Google’s AI system that beats professional human players in the game of Go] is seen as the peak in AI. However, there are rules governing the game of Go, and therefore all the moves are calculable,” Yu Wang, a director at China Automotive Technology and Research Center, said.
“In reality, there are no rules governing roads and weather conditions," Wang told FinanceAsia. "I think it will take over 10 years for AI to be mature enough to reach that level [of handling various road conditions]."
