Japan’s birth rate is at its lowest level since records began in 1899. It is estimated that, by 2036, one in three people in the country will be elderly. This decline in population has led to the emergence of fresh challenges, and technology is well placed to serve them.
This is an extract from a white paper published by Japan’s Ministry of Land, Infrastructure, Transport and Tourism: “With regard to the automotive transportation industry, which includes trucks, buses and taxis, drivers work in an environment in which annual work hours are 10% to 20% longer, and annual wages are 10% to 30% lower when compared to the average of all professions, so the lack of drivers is even more ingrained and securing workers is a pressing issue.”
They also described the precarious situation with respect to drivers in the country’s logistics sector where shortages are now “especially evident” and the sector has ageing truck drivers and “possibilities of increased difficulties” in securing personnel in the medium to long term.
It is amid this backdrop that Japan is toying with self-driving vehicles. In an article published by Transport Topics, President of Hinomaru Kotsu, a major taxi operator in the country said, “We would like to introduce self-driving cars to tackle taxi driver shortages.”
The publication also reported that two companies are already conducting tests in an area between the Otemachi and Roppongi districts of Tokyo.
The Hype is Quieting, but Progress Isn’t
There is plenty of news about self-driving cars being the transport of the future, but is there more hype than actual progress? Can these cars handle the unpredictable situations that humans deal with every day when they are driving?
In order to explore where we are with self-driving vehicles, Binary District Journal spoke with Matti Kutila, Senior Scientist at VTT, a company in Finland that has devised a car, called Marilyn, which can reportedly see better in a fog than humans can.
“The development will take steps forward and self-driving will be a reality in 2030-2040”
“I wouldn’t say that development and interest has disappeared in our key customer domains,” Matti says. “However, you are right that we have seen that self-driving become a less sexy topic in the media compared to last year.”
There is also the question of whether perceptions around the self-driving segment have been less than realistic. “This is not a bad thing since expectations last three years were sometimes unrealistic,” Matti adds.
“However, the development will take steps forward and self-driving will be a reality in 2030-2040 - not necessarily 24/7 but most of the time in any case.”
Development is Ramping Up
When it comes to self-driving vehicles, there are different levels of autonomy. There are already cars with Advanced Driver Assistance Systems (ADAS) in place.
These systems are designed to help humans make fewer errors and reduce accidents. Their scope extends to adaptive cruise control, collision avoidance, traffic and sat nav warnings, lane departure warning systems, etc.
“The automotive AI market was valued at $782.9 million in 2017 and is expected to reach $10.6 billion by 2025”
These systems are driven by Artificial Intelligence (AI), which is playing a central role. So much so that, according to a report, the automotive AI market was valued at $782.9 million in 2017 and is expected to reach $10.6 billion by 2025 – CAGR of 38.46%.
Developments are already on the way that will add exciting features to self-driving cars and increase their capabilities.
VTT’s Marilyn platform, for example, follows waypoints that connect to the city infrastructure and allow the vehicle to capture free parking spaces. The cars also have ‘modern environment perception’ devices which allow the vehicles to be perceptive about their surroundings and interact with each other.
Tech Can See What We Can’t
Technology can sometimes provide humans with an edge. A simple invention like the microscope that we take for granted has revolutionised medicine.
The machine enables us to see things with our eyes that nature doesn’t allow for. When it comes to self-driving vehicles, a similar edge could improve human skills or eliminate the need for them.
VTT deploys LiDAR (Light Detection and Ranging), which uses lasers much in a similar way to the radio waves used in Radar. Matti informed us that a number of startups in the United States are using this technology to develop better resolution.
However, as far as vehicles are concerned, in-vehicle satellite positioning and inertia units are required to conduct automation and control it. There are also limitations to LiDAR such as cold weather and fog.
Yet, VTT’s car can be driven in adverse conditions, albeit slowly, compared with other self-driving vehicles that may come to a complete stop. In fact, in simulated fog tests, the car can drive itself where humans fail to see anything.
Mistakes Are Going to Happen
Driving can provoke strong emotions. Road rage is not unheard and it would not be ridiculous to expect people to get similarly worked up with driverless cars.
In March 2018, an Uber self-driving car killed a woman in Tempe, Arizona. Uber was not charged with any criminal proceedings. The company has begun retesting its cars in Pittsburgh in a limited fashion. In Arizona, Waymo, a project of Google, has also been testing its own self-driving vehicles.
“It’s a bit like after an airplane crash, people start thinking about aviation safety but then forget about it after a while”
These cars have been eliciting strong reactions from the locals, according to CNBC, who are frustrated with the conservative driving of the vehicles, especially at intersections. People have been driving around them and in some cases pointing guns at the cars.
In Chandler, Arizona where these tests are taking place, there have been cases of slashed tires, rocks thrown and vandalism.
It may appear on the surface that people are enraged about the car’s abilities or inabilities. The idea that these cars will take a crucial job away from humans may make some people angry, while others may be unhappy with their current limitations and the testing process causing them real or perceived inconveniences.
“Most of the people don’t like that much driving in monotony highways and they rather would give control to the vehicle itself,” Matti says. However, he does acknowledge that tragedies like the one in Arizona could evoke strong reactions.
“If something catastrophic happens, attitude of people may change suddenly,” he says. “It’s a bit like after an airplane crash – people start thinking about aviation safety but then forget about it after a while.“
Overall, though, he feels that the self-driving cars have an edge. “When people and especially young generation get used to self-driving, I don’t think that they will go back to manual driving.”
When we consider the possibilities presented by driverless cars, the smaller details are often forgotten. Take insurance, for example. The logistics of having two separate insurance premiums for the car when it is in manual or automated mode respectively is prohibitively complex. Does a premium change when an AI is in control of a car that has the option to control it manually installed as well? Insurers have been made to compromise.
Many automated cars will come under one insurance policy that covers drivers both when they are manually in control of the vehicle and when they are being driven by the AI itself. Under this proposed system, it could be up to insurance companies to recoup some of the money they pay out from the car manufacturers themselves. Ultimately, an influx of driverless cars will lead to a drastic reduction in the number of accidents seen on the road, despite any fears around their safety on a case-by-case basis, so insurers should find themselves paying out less often regardless.
Illustrations by Kseniya Forbender
To contact the editor responsible for this story:
Margarita Khartanovich at [email protected]
- How Vulnerable Is the Internet of Things (IoT) to Hackers?
- Blockchain Interoperability Remains a Critical Missing Puzzle Piece
- How to Keep A Business Safe: Tips From IBM and Shapeshift.io Cybersecurity Experts
- China Vs the US: Who Will Win the AI Race?
- AI Applications and “Black Boxes”: How to Make Use of Recent Research on Artificial Neural Networks