Tesla Autopilot 2.0 to feature three cameras and more radar sensors

Tesla's driverless car system, Autopilot, looks set to get a major upgrade.

According to reports by Electrek, Tesla is testing out new hardware that will see the driverless technology use three cameras at the front of the car as well as radar sensors dotted around the vehicle.

The radar sensors will be placed at the corners of the car to mimic lidar and give the car a sense of its surroundings.

The report said that Tesla is installing the new system housing into its cars. The hardware would initially use current Autopilot software but would add new features and improvements over time.

The report said that while Autopilot uses live data from its sensors, it also relies a lot on GPS data and "high-precision maps" built up from data collected across its fleet of cars.

It is not known when Autopilot will be launched fully, but Tesla owner Elon Musk hinted in an earnings call that technology is on the way that will "will blow people's minds".

05/08/2016: Google driverless car in hit and run

One of Google's self-driving cars has been involved in a hit-and-run incident, sustaining minor damage.

The collision occurred in Los Altos, California on the 15th of July, and was revealed by a report from California's Department of Motor Vehicles.

The driverless car was waiting at a stop sign when an unknown driver shunted the back of it at 7mph, causing "minor damage to its rear hatch and sensor".

"There were no injuries reported at the scene," the report stated. However, it noted that "Google was unable to determine whether or not the other vehicle sustained any damage," due to the fact that "the driver of the vehicle left the scene without exchanging information".

The incident comes hot on the heels of another self-driving car crash, in which the driver of a Tesla Model S was killed. The car was over the speed limit, and had the 'Autopilot' mode engaged.

While Google's autonomous vehicles have previously been involved in collisions, their track record is quite good. One of the few occasions where the self-driving cars have been found at fault was an incident in February, where one of the company's fleet crashed into a public bus in Mountain View.

30/06/2016: Intel, MobilEye and BMW 'team up on driverless cars'

Intel is said to be working on plans for driverless cars with BMW and Israeli firm Mobileye.

The trio of companies will hold an event in Munich, Germany this Friday to talk about the driverless car initiative, according to Bloomberg. It is not known how extensive the partnership will be or what form it will take.

Mobileye has provided cameras, components and software for driverless cars to General Motors and Tesla, and BMW has also been a customer of the firm.

Intel is looking to supply the processors needed to crunch sensor data in order to make decisions on driving. The chip firm wants to get involved in autonomous cars as the PC market continues its decline, while Intel also missed the mobile revolution.

BMW CEO Harald Krueger in May confirmed his firm would produce a driverless car dubbed the 'i Next' in 2012 after its new i8 Roadster in 2018.

"This will be followed in 2021 by the BMW i Next, our new innovation driver, with autonomous driving, digital connectivity, intelligent lightweight design, a totally new interior and ultimately bringing the next generation of electro-mobility to the road," he said.

Reuters corroborated the story and said that sources "familiar with the matter" claimed the three firms are close to unveiling the partnership.

24/06/2016: Autonomous cars should sacrifice passengers to save pedestrians

Should driverless cars prioritise the lives of their passengers above all else?

Most people don't think so, according to a recent study conducted by the Massachusetts Institute of Technology. The surveys revealed that 76 percent of respondents believed it would be more ethical for an autonomous vehicle to sacrifice one passenger if doing so would save the lives of 10 pedestrians.

This was just one scenario that the study presented. Researchers introduced variations to the example to make the decision more difficult, such as suggesting a family member or a child was in the car, according to Forbes, but most participants still believed the car should save the most people possible.

Eighty-one per cent of participants, however, said they would not buy a car that had been programmed this way.

"Most people want to live in in a world where cars will minimise casualties," says Iyad Rahwan, one of the authors of the study told PA. "But everybody wants their own car to protect them at all costs."

It's not the first time studies have found that people are a bit wary of owning a driverless car. Recent surveys have found that 65 per cent of British motorists believe a human should always be in control of a vehicle and that 70 per cent would not feel confident being a passenger in an autonomous car.

Fears about cars making moral decisions are a bit moot at the moment, the Inquirer points out, considering all accidents involving Google's self-driving cars have been minor and almost always the fault of a human. It is something that will need to be taken into consideration at some point, though, with Business Insider predicting that 10 million self-driving cars will be on the streets by 2020.

"Life-and-death trade-offs are unpleasant, and no matter which ethical principles autonomous vehicles adopt, they will be open to compelling criticisms," said Professor Joshua Greene of Harvard University told Science. "Manufacturers of utilitarian cars will be criticised for their willingness to kill their own passengers. Manufacturers of cars that privilege their own passengers will be criticised for devaluing the lives of others and their willingness to cause additional deaths. "

17/06/2016: Rolls-Royce shows off driverless concept car

Rolls-Royce has unveiled what its take on a luxury driverless car could look like.

The new EX103 Vision Next 100 concept car is touted as "purely visionary", meaning that such a car is unlikely to ever get made. It was demonstrated at a London event celebrating Rolls-Royce owner BMW's centenary.

The car is 5.9 metres long, about the same length as the company's Phantom. A single door opens up to reveal a cream-coloured interior, which the firm describes as "a beautifully textured, ivory-coloured luxurious throne upon which our passengers are conveyed, and from which they command".

Inside, an HD screen sits at the front of the cockpit instead of a driving wheel and controls. The car is under the control of a digital AI assistant called Eleanor, which takes orders from passengers. The assistant is named after the actress Eleanor Thornton, who is thought to have inspired the car's Spirit of Ecstasy bonnet ornament.

"She brings the car around when her passengers are ready to travel and, whilst conveying her charges to their next destination, helpfully but discreetly makes suggestions and recommendations, briefing them ahead of their arrival so they are ready to perform," the company said in a statement.

The roof of the car is described as a "canopy over the occupants that provides privacy while allowing them to contemplate the majesty of the stars in the firmament above as they glide through the night".

Outside, all the wheels are almost completely covered to the point where we'd worry how it would ever tackle speed ramps.

The car is billed as a zero-emission vehicle that "rejects the notion of anonymous, utilitarian and bland future modes of mobility", although the company did not say how the car would be powered.

Torsten Mller-tvs, chief executive officer at Rolls-Royce Motor Cars, said the arrival of the Rolls-Royce Vision Next 100 "boldly points to a bright future for our marque where our patrons' individual demands for complete and authentic personalisation will be met through an exquisite fusion of technology, design and hallmark Rolls-Royce craftsmanship."

Picture credit: Alphr

03/06/2016: Driverless cars: Google is teaching autonomous cars when to honk at other drivers

Google's self-driving cars are being taught how to honk.

The company's engineers have been training their autonomous vehicles to use their horns, but, according to Google's Self-Driving Car Project Monthly Report, only when it is "polite, considerate and ... makes driving safer for everyone." In other words, the cars are not only being taught to honk but are also being trained to know when it is appropriate to do so. If one of the driverless cars were to detect an impending collision, for example, it would know to use its horn as an alert.

During testing, the vehicles were trained to distinguish between potentially dangerous situations and false positives. For example, they would know to honk if another car was driving down the wrong side of the road, but not if a car was facing the wrong way during a three-point turn.

"At first, we only played the horn inside the vehicle so we wouldn't confuse others on the road with a wayward beep," the report stated. "Each time our cars sound the horn, our test drivers take note whether the beep was appropriate, and this feedback helps our engineering team refine our software further."

Additionally, the vehicles are being taught to use different types of honks for different situations. If another car were slowly backing up toward one of the self-driving vehicles, it might let out "two short, quieter pips" to let the driver know it's there. In more urgent situations, however, the vehicle would use "one loud sustained honk."

Google itself admits that given the amount of time its vehicles spend on busy streets, it is inevitable they will be involved in collisions. In fact, they already have. In February, a Google Lexus autonomous SUV drove itself into a bus in Mountain View, California.

According to Gizmag, Google's self-driving cars had been involved in at least 17 accidents before that, but this marked the first time the company conceded that its vehicle bore some responsibility for the collision. Hopefully, teaching the cars to use their horns will help prevent something like this from happening again.

"Our goal is to teach our cars to honk like a patient, seasoned driver," said Google in the report. "As we become more experienced honkers, we hope our cars will also be able to predict how other drivers respond to a beep in different situations."

19/05/2016: Google patents sticky cars to protect pedestrians from driverless vehicles

Google has patented a way of minimising accidents involving driverless cars by glueing people they hit onto the bonnet.

The patent was filed in 2014 and designed as a temporary measure to keep people safe while driverless technology improves.

The US Patent and Trademark Office awarded the patent that sees cars adding an adhesive layer to the bonnet of the car in order to reduce damage if hit.

The patent description said: "Ideally, the adhesive coating on the front portion of the vehicle may be activated on contact and will be able to adhere to the pedestrian nearly instantaneously.

"This instantaneous or nearly instantaneous action may help to constrain the movement of the pedestrian, who may be carried on the front end of the vehicle until the driver of the vehicle (or the vehicle itself in the case of an autonomous vehicle) reacts to the incident and applies the brakes.

"As such, both the vehicle and pedestrian may come to a more gradual stop than if the pedestrian bounces off the vehicle."

The patent added that "existing technology found in production vehicles does little to mitigate the secondary impact a pedestrian may experience".

While the patent can apply to any vehicle, Google said that it has been designed with its autonomous vehicles in mind. The sticky coating is described by Google as lying under an outer eggshell layer that would break in the event of an impact. This means that the car doesn't drive around like a mobile piece of flypaper scooping up insects as it travels.

It is not known if Google plans to use the technology on its driverless vehicles in the future. A Google spokesperson told San Jose Mercury News: "We hold patents on a variety of ideas. Some of those ideas later mature into real products and services, some don't."

Rene Millman

Rene Millman is a freelance writer and broadcaster who covers cybersecurity, AI, IoT, and the cloud. He also works as a contributing analyst at GigaOm and has previously worked as an analyst for Gartner covering the infrastructure market. He has made numerous television appearances to give his views and expertise on technology trends and companies that affect and shape our lives. You can follow Rene Millman on Twitter.