Getting To The Self-Driving Car

There is still much work to be done, but that hasn’t dampened the buzz around it.


Realizing the vision of the fully autonomous vehicle is one of the most ambitious research and development initiatives since the Apollo program of the Space Age.

While the goal of Apollo was to send a man to the Moon and safely return him to Earth, the goal of self-driving cars is to get a person out from behind the steering wheel and safely convey that person to home, work, a vacation resort, and other destinations.

It is worth noting that Baidu has dubbed its autonomous vehicle program, aimed at developing a global ecosystem for automated driving technology, as Apollo. That program has attracted the participation of more than 100 companies around the world since it kicked off in April of last year. Baidu marked the first anniversary of its launch with the introduction of Apollo 2.5, the latest version of its open autonomous driving platform.

The Chinese company has offered up more than 200,000 lines of code, shared with more than 2,000 Apollo developers and partners. Baidu is working with Beijing Electric Vehicle, Chery Automobile, and China FAW Group to establish joint R&D laboratories in China. The effort also takes in five of China’s top universities.

Qi Lu, the chief operating officer of Baidu, said the Apollo platform aspires to be the “Android of the autonomous driving industry, but more open and powerful.”

Smart transportation, or whatever you want to call it, is a topic of interest at nearly every high-tech conference these days, including the Design Automation Conference and SEMICON West.

The annual Internet of Things World conference, held in May at the Santa Clara Convention Center in Silicon Valley, ran concurrently with a two-day program, Connected & Autonomous Vehicles, and there were state-of-the-art vehicles on the IoT World exhibit floor. The vehicle conference featured sessions on the topics of “Mobility as a User Enterprise,” “Integrating Cyber-Security from the Ground Up,” “The Global Nature of Autonomous Driving Innovation,” “Driver Assisted – Insurance Reduced Connected Cars,” and “Automakers, the New Service Providers.”

Also staged in May was NIWeek 2018, put on by National Instruments in Austin, Texas. The company devoted one full keynote session to “The Future of the Automotive Industry.” There also were technology breakout sessions on related subjects, such as artificial intelligence, machine learning, big data, sensor fusion, and automotive connectivity.

Tomohiko Adachi, senior principal engineer at Mazda Motor, spoke at a morning keynote session during NIWeek.

”Mazda’s electronics testing and research group develops logic and robustness for each electronics component. Logic refers to the functionality of each component, and robustness refers to the component’s ability to operate. It is very difficult to find the system that the technology needs, so we built our own. And we knew it would be large and complex,” Adachi said. “Mazda is focused on advancing the development and practical use of a model-based solution to stay ahead of our competition. With this hardware innovation, it is only natural that, where possible, we wanted to use models to gauge performance. However, certain components are very difficult to model. And it is very challenging to buy these systems where we have interaction between the human driver and the vehicle.”

He further described how Mazda was able to achieve a 90% reduction in testing time, while improving test coverage using better test instrumentation. Adachi noted that he was inspired to work on advanced automotive electronics by the “Knight Rider” television series of the 1980s and its KITT supercar with artificial intelligence.

No driver needed
Autonomous driving has captured the tech world’s imagination.

“It’s a pretty exciting time for the automotive market,” said Kamal Khouri, vice president and general manager of advanced driver-assistance systems at NXP Semiconductors. “We’re actually experiencing three revolutions at the same time. I want to dive a little deeper into the topic of autonomy and autonomous driving – specifically, talk about our responsibilities around safety and security, so that we deliver these products to our families, to our customers, and they’re going to work the way we expect them to.”

Sensing, thinking, and acting are key to autonomous driving, according to Khouri. “Once we develop that view, that dynamic view around the vehicle, we have to decide what we want to do based on our objective. We’re trying to get from point A to point B, and we have all these things happening around us, and we need to make a decision. We have to think about these things and make a decision. And that decision is critical and has to be correct, because once you make that decision, you need to act on it, you have to execute on it. All of those mechanisms get the car to drive, steer, and stop,” he said.

He continued, “All of this is going to require immense amounts of computing. This computing is going to be opposing itself. This is in the realm of machine learning, and AI, and neural networks. This involves a lot of probabilistic computing.”

Khouri noted that environmental safety will be an essential component of autonomous vehicles in addition to functional safety. So will security and privacy, because valuable data will be flowing in and out of a vehicle, including payment information.

“We must protect privacy,” Khouri asserted. “We have to prevent unauthorized access. This is not just about stealing personal data. Remember, these vehicles on wheels can be used as weapons. We’ve seen this, unfortunately. We’ve seen this happen. How do we make sure that malicious individuals do not get control and ahold of these vehicles and cause havoc? By doing so, we’re increasing the safety of autonomous driving, and saving lives, as well, with the introduction of this technology. And last but not least, we still have a promise to our customers to deliver the automotive-grade solutions that they’ve gotten used to. We talked about safety, we talked about security, and then there’s the automotive reliability. When we go to a dealership and buy a car, we’re not taking it back and fixing problems with it and having to deal with issues. That is the promise that we have to give our customers.”

Bryant Walker Smith, an assistant professor at the University of South Carolina’s School of Law (he also teaches at South Carolina’s School of Engineering), said autonomous vehicles will require technical genius, but it also will require humility and responsbility. He looked back at a transportation technology that electrified the United States in the late 19th century – the “steel horse,” aka the bicycle. “It is on the front page of newspapers,” he noted. “The bicycle is the talk of the town in the late 1800s.”

Automobiles, the first human flight, and AM radio eventually eclipsed attention paid to the bicycle, said Smith. Horses deposited much manure on city streets and also posed the issue of dying on those streets.

“The internal combustion engine came along, and people said, ‘That is the green transportation mode. We have solved our urban pollution problem forever.’ Until we realized that maybe we hadn’t,” Smith said. “In 1911, about 10% of the Oldsmobiles sold would just catch on fire. You’d be driving along, you’d look back and say, ‘Oh, my car’s on fire; I have to put that out again.’ You’d then put out the fire and carry on your way.”

He quoted entrepreneur Elon Musk, who said that everything is beta. “We should look at what existing law requires and decide whether we need more permissiveness or more restrictiveness,” he noted. “As developers, we have to ask, ‘Do we ask for permission or do we ask for forgiveness?’ Engineers solve problems. But, engineers cause problems, too. That’s the story of innovation, of engineering, of law, of public policy, that we take one set of problems, we replace that set with a new set of problems, and we just really, really hope that our new set in aggregate is smaller than our old set. That will be true with automated driving, as well. And it’s something we always have to recognize, to be honest about the tradeoffs.”

In the future, many products will be more like services, which will present a regulatory issue, according to Smith. “The key is trustworthiness. The trustworthy company keeps its promises. It shares its successes and its failures. That’s where a company earns its credibility.”

Richard Aspinall, manager of powertrain manufacturing at Faraday Future, an electric vehicle startup, was bullish on the automotive market, as well, and for good reason—the company recently received $2 billion in private funding from an overseas investor. The transaction was cleared by the federal Committee on Foreign Investment in the United States.

Fig. 1: Faraday Future’s new SUV. Source: Faraday Future.

Aspinall disclosed some details on the company’s new electric SUV, which he said can reach speeds of up to 150 miles per hour.

In the startup ecosystem of Los Angeles, Faraday has exchanged lawsuits with other companies over poaching of employees in its “fight to the first,” Aspinall said.

There is plenty of buzz surrounding autonomous vehicles. But there also are a number of issues that are unresolved, including government regulation, insurance liability concerns, and testing practices. And those are in addition to the research and development of the technology, which has a long way to go to proceed beyond Level 3 autonomy.

Leave a Reply

(Note: This name will be displayed publicly)