Driverless Cars: Uber car involved in fatal crash had software flaws

07/08/2017: Government draws up cyber security proposals

The UK government published tougher cyber security guidance for car manufacturers yesterday to ensure internet-connected vehicles will be better protected from hackers.

The proposals state that manufacturers should create and adhere to a security programme tied to their "broader mission and objectives", with the Department for Transport (DfT) worried that the cars will be targeted by hackers to access personal data, steal cars that use keyless entry, or even try to take over control of the cars.

Eight basic principles will form the basis of new legislation to be put before Parliament, the DfT said. Firstly organisational security must be a board-level priority, with the boardroom taking responsibility for it. Engineers must also be aware of current threats and work with third-parties to mitigate them as they build smart vehicles, the proposals state.

Manufacturers must also offer product aftercare and incident response to ensure systems are secure over their lifetime, not just upon their release, as well as designing the systems using a defence-in-depth approach, which means that cloud servers to which any vehicle data is sent are independently protected, as well as there are no single points of failure.

Firms must also work with their sub-contractors, suppliers and wider supply chain to co-ordinate on security, to agree standards and data requirements so that their security systems work together. Lastly, the systems themselves should be made resilient to cyber attacks like sensor jamming, and are fail-safe in the event of a successful hack.

Transport minister Lord Callanan said: "Our cars are becoming smarter and self-driving technology will revolutionise the way in which we travel. Risks of people hacking into the technology might be low, but we must make sure the public is protected."

He added: "Whether we're turning vehicles into wifi connected hotspots or equipping them with millions of lines of code to become fully automated, it is important that they are protected against cyber attacks."

Mike Hawes, the CEO of the Society of Motor Manufacturers and Traders, welcomed the measures, saying it's vital to champion cyber security and share best practice. He added: "These vehicles will transform our roads and society, dramatically reducing accidents and saving thousands of lives."

Driverless car technology has been the focus of many companies and there is plenty of competition, from Uber to Baidu.

The government is also looking at creating a new framework to insure self-driving cars, announced in this year's Queen's speech under the Autonomous and Electric Vehicles Bill.

02/08/2017: Tim Cook says Apple AI is applicable to more than just cars

Apple's research into autonomous car technology might be applicable to sectors other than the automotive industry, according to CEO Tim Cook.

The tech giant finally acknowledged the truth in rumours that it was building driverless technology in June, when Cook told Bloomberg that it was "a core technology that we view as very important".

But he declined to give a steer on how the tech would manifest itself in Apple products. Yesterday he painted a clearer picture of its potential on a conference call following the company's quarterly results.

"Autonomy is the mother of all AI projects," he said, as transcribed by Seeking Alpha. "And the autonomous systems can be used in a variety of ways and a vehicle is only one. But there are many different areas of it and I don't want to go any further with that."

While Cook is keeping quiet on Apple's plans for autonomous tech, it's clear that it is not limited to vehicle software. If it's the "mother of all AI", it's possible that its applications could expand into various different areas.

Apple already has a smart voice assistant in Siri, and its rivals haven't been slow to deploy AI in chatbots, home speakers, and much more. Ocado is using artificial intelligence to automate warehouses, for instance, while CIOs believe AI could help them automate plenty of routine business functions.

28/06/2017: Self-driving cars are failing to cope with the sight of a kangaroo in front of them.

According to a report by ABC News in Australia, the means by which the marsupials move bamboozles autonomous cars.

This was noticed by Volvo Australia's technical manager David Pickett, who said that when a kangaroo is in mid-flight it looks further away, and then when it lands it looks closer. As self-driving cars use the ground as a reference point, a bounding kangaroo confuses the car and thus cannot determine its distance.

But Pickett told ABC News that it gets even more complicated; the car needs to identify a kangaroo.

"We identify what a human looks like by how a human walks, because it's not only the one type of human you've got short people, tall people, people wearing coats. The same applies to a roo," he told reporters.

While Volvo has tested its large animal detection software in moose in Sweden, solving the kangaroo problem is no nearer to a solution. The country's National Roads and Motorists' Association said that 80% of animal collisions in the country involve kangaroos.

While this wouldn't delay the rollout of driverless cars in Australia, the problem still needed to be sorted out before introduction.

Kangaroos aren't the only issue for self-driving cars in the Lucky Country. The continent also features unsealed roads, unmarked highways and massive road trains. But generally, the feeling is that although while road infrastructure and technical issues need sorting out, Australia will still be on course to have cars on the road before many other places.

"The maturity is much further along than maybe is publicly aware," said Australian Driverless Vehicle Initiative executive director Rita Excell.

27/06/2017: Nvidia signs AI deal

Nvidia has announced partnerships with several car makers and automotive suppliers to boost artificial intelligence in driverless cars.

Volkswagen will cooperate with Nvidia to expand its research on deep learning. At the carmaker's Data Lab, researchers are working with Nvidia on developing new procedures for optimising traffic flow in cities.

The firms have established a startup support programme at Volkswagen's Data Lab to provide technical and financial support for international startups developing machine learning and deep learning applications for the automotive industry. The pair will accept five startups to the support programme from this autumn.

There will also be a "Summer of Code" camp where high-performing students with qualifications in IT, mathematics or physics will have an opportunity to develop deep learning methods in teams and to implement them in a robotics environment.

Nvidia will also work with Volvo and Autoliv to develop advanced systems and software for AI self-driving cars. The three companies will work together along with Zenuity, a newly formed automotive software development joint venture equally owned by Volvo Cars and Autoliv, to develop next-generation self-driving car technologies.

Production vehicles built on the Nvidia Drive PX car computing platform are planned for sale by 2021.

Volvo Cars, Autoliv and Zenuity will use Nvidia's AI car computing platform as the foundation for their own advanced software development. The platform uses cameras and GPS to plan a safe route and adjust to changing circumstances.

Zenuity will also provide Volvo with self-driving software. Autoliv will sell this software to third-party OEMs using its established and broad sales, marketing and distribution network.

Nvidia has partnered with automotive supplier ZF and camera perception software supplier Hella to deploy AI technology on the New Car Assessment Program (NCAP) safety certification for the mass deployment of self-driving vehicles.

ZF and Hella will provide customers with a self-driving system that integrates front camera units, as well as supporting software functions and radar systems.

The firms will use Nvidia's Drive AI platform to develop software for scalable modern driver assistance systems that connect their advanced imaging and radar sensor technologies to autonomous driving functionality.