At the blockbuster plenary sessions, the chairs stretched so far back that even the most youthful Silicon Valley college dropouts-turned VC hoovers had to squint to see the action up in front. A handful of large projection screens hung between the ballroom's chandeliers, displaying loop-de-looping flow charts on vehicle safety systems, sensor alignments, liability law. But despite the best efforts of the downtown San Francisco Hilton's air conditioners, the air shared by the attendees of this year's Automated Vehicles Symposium was thick with secrets and doubt. Eight years after Google first showed its self-driving car to The New York Times, the autonomous vehicle industry is still trying to figure out how to talk about itself. Over the three-day conference, engineers, business buffs, urban planners, government officials, and transportation researchers grappled with how to tell the public that its wonder drug of a transportation solution will have its limitations.
Americans spend 8 billion hours stuck in traffic every year. Deep neural networks can help! DeepTraffic is a deep reinforcement learning competition. The goal is to create a neural network to drive a vehicle (or multiple vehicles) as fast as possible through dense highway traffic. What you see above is all you need to succeed in this competition.
They may resemble giant metal toasters on wheels, but these strange-looking vehicles could one day save you a trip to the supermarket. The American grocery store chain Kroger is teaming up with Nuro, a Silicon Valley-based robotics company, to test a fleet of robotic cars this fall in a yet-to-be-announced city. The new autonomous technology is designed to meet changing demands in the grocery market. "Our customers are increasingly wanting different ways of fulfilling their food and shopping needs," said Yael Cosset, Kroger's chief digital officer. The deliveries will be carried out by Nuro's R1 car bots, which have a top speed of 25 miles per hour, stand 6 feet tall and measure about "half the width of a Toyota Corolla," Nuro CEO Jiajun Zhu told NBC News MACH in an email.
Traffic hell is alive and well in Los Angeles. In 2017, Angelenos were stuck on the road for 102 hours each (more than four full days), costing the city $19.2 billion, according to INRIX's annual global traffic scorecard. Traffic is almost as bad--and costly--in Moscow, Sao Paulo, and London. But this is the 21st century! Can't AI fix these problems by optimizing traffic flow?
Australian startup Baraja is giving car manufacturers an alternative to the 13kg drum Uber is using for driverless vehicle mapping, launching its Light Detection and Ranging-based solution to help progress driverless vehicles. The product, Spectrum-Scan, uses shifting wavelengths of light to create "eyes" for autonomous vehicles. CEO and co-founder Federico Collarte told ZDNet the lidar solution solves the scalability, reliability, and performance issues that have "challenged automakers, rideshares, and the tech behemoths as they race toward a fully-autonomous future". At this point, it doesn't matter what type of vehicle the sensors are mounted to, as Collarte said "today you don't buy self-driving cars, you buy cars and you give them self-driving capabilities". Up to four sensor heads are connected via fibre optic cable to a central processor.
In what is potentially a world-first, researchers have launched a pilot program which replaces physical traffic light systems with virtual alternatives. Traffic light systems often play a crucial role in traffic flow. When road systems are organized effectively, they can provide a way to increase the efficiency of travel. However, in poorly-managed setups, traffic lights may leave drivers cursing at the wheel behind a red light on a deserted road or may also cause huge traffic jams. Love or hate them, for many road systems worldwide, they are necessary -- but this does not mean they cannot be improved.
Uber's dreams of a fleet of self-driving taxis may be on the rocks, if the firm's latest move is anything to go by. The ride-hailing company laid off 100 safety drivers after autonomous vehicle tests were suspended in the US, following a high profile crash in Arizona. Uber initially said it was not shuttering its entire autonomous vehicle program in the aftermath of the incident, in which 49 year old Elaine Herzberg died. Instead, it announced it was focusing on more limited testing in Pittsburgh, Pennsylvania and California, aiming to resume self-driving this summer. That decision may have been revised, if the latest news is anything to go by, with all 100 redundancies at its Pittsburgh base of operations.
Mercedes Benz owner Daimler is teaming up with Bosch to launch a fleet of driverless taxis in California's Silicon Valley next year. It is part of a program to test vehicles designed for city driving in an attempt to keep up with the likes of Waymo and Uber. The world's largest maker of premium cars and biggest automotive supplier gave few details about their robo-taxi program, described as a passenger shuttle service, and did not reveal which city would host it. Mercedes boss Daimler is teaming up with Bosch to launch a fleet of driverless taxis in California's Silicon Valley next year. Negotiations with the municipality within the sprawling technology hub of Silicon Valley were still underway, spokespersons for the companies said on a conference call with journalists.
Federal prosecutors have charged a former Apple employee with stealing trade secrets related to Apple's autonomous vehicle program. Xiaolang Zhang allegedly worked on Apple's secretive self-driving car project. Zhang left Apple in April saying he was going to work for a Chinese electric vehicle company called Xpeng Motors. He is accused of copying more than 40GB of Apple intellectual property to his wife's laptop before leaving the company, according to court documents. The documents do not accuse Xpeng Motors of wrongdoing.
Like in a Tough Mudder, you've got a few strategies when it comes to the race to launch a taxi-like service with autonomous vehicles. You can start early and keep a slow but steady pace. You can show up a bit late, then try to sprint through it. Or you can hold back, see what trips up other contenders, and then slowly work your way through the obstacles. The big automakers tend to fall into the third category.