Incident 25: Near-miss between two Self-Driving Cars

Description: A Google self-driving car allegedly cut off a Delphi self-driving car during a road test, however the Delphi car sensed and avoided collision with the Google car.
Alleged: Google developed and deployed an AI system, which harmed Delphi Technologies.

Suggested citation format

Anonymous. (2015-05-11) Incident Number 25. in McGregor, S. (ed.) Artificial Intelligence Incident Database. Responsible AI Collaborative.

Incident Stats

Incident ID
25
Report Count
11
Incident Date
2015-05-11
Editors
Sean McGregor

Tools

New ReportNew ReportNew ResponseNew ResponseDiscoverDiscover

CSET Taxonomy Classifications

Taxonomy Details

Full Description

A Google self-driving car cut off a Delphi self-driving car in Silicon Valley, California. A Delphi spokesperson first reported the near miss under the context of both cars acting in the way they should. The Delphi car sensed the Google car's approach into the lane it intended to merge into, therefore termintating the Delphi car's lane change until safe to do so. Google agreed with this statement. Delphi later amended their statement to say "the vehicles didn't even come that close to each other."

Short Description

A Google self-driving car allegedly cut off a Delphi self-driving car during a road test, however the Delphi car sensed and avoided collision with the Google car.

Severity

Negligible

AI System Description

Self-driving cars developed by Google and Delphi, respectively

System Developer

Google, Delphi

Sector of Deployment

Transportation and storage

Relevant AI functions

Perception, Cognition, Action

AI Techniques

environmental sensing, decision trees, artificially intelligent automobiles

AI Applications

autonomous vehicles, interpreting traffic patterns

Location

Silicon Valley, California

Named Entities

Google, Delphi

Technology Purveyor

Google, Delphi

Beginning Date

6-2015

Ending Date

6-2015

Near Miss

Near miss

Intent

Accident

Lives Lost

No

Infrastructure Sectors

Transportation

Data Inputs

traffic patterns, environmental input

Incident Reports

Four of the nearly 50 self-driving cars now rolling around California have gotten into accidents since September, when the state began issuing permits for companies to test them on public roads. Scott Budman reports. (Published Monday, May 11, 2015)

Four of the nearly 50 self-driving cars now rolling around California have gotten into accidents since September, when the state began issuing permits for companies to test them on public roads.

Two accidents happened while the cars were in control; in the other two, the person who still must be behind the wheel was driving, a person familiar with the accident reports told The Associated Press.

Three involved Lexus SUVs that Google Inc. outfitted with sensors and computing power in its aggressive effort to develop "autonomous driving," a goal the tech giant shares with traditional automakers. The parts supplier Delphi Automotive had the other accident with one of its two test vehicles.

Google and Delphi said their cars were not at fault in any accidents, which the companies said were minor.

Since September, any accident must be reported to the state Department of Motor Vehicles. The agency said there have been four, but would not comment about fault or anything else, citing California law that collision reports are confidential.

The person familiar with the accident reports said the cars were in self-driving mode in two of the four accidents, all of which involved speeds of less than 10 mph. The person spoke on condition of anonymity because they were not authorized to discuss the reports publicly.

Five other companies have testing permits. In response to questions from the AP, all said they had no accidents. In all, 48 cars are licensed to test on public roads.

The fact that neither the companies nor the state have revealed the accidents troubles some who say the public should have information to monitor the rollout of technology that its own developers acknowledge is imperfect.

John Simpson, a longtime critic of Google as privacy project director of the nonprofit Consumer Watchdog, pointed out that the company's ultimate goal is a car without a steering wheel or pedals. That would mean a person has no power to intervene if a car lost control, making it "even more important that the details of any accidents be made public — so people know what the heck's going on."

A chief selling point for self-driving cars is safety. Their cameras, radar and laser sensors give them a far more detailed understanding of their surroundings than humans have. Their reaction times also should be faster. Cars could be programmed to adjust if they sense a crash coming — move a few feet, tighten the seat belts, honk the horn or flash the lights in hope of alerting a distracted driver.

A higher priority so far is teaching them to avoid causing a serious accident that could set public and political acceptance of the technology back years, said Raj Rajkumar, a pioneer of the technology with Carnegie Mellon University.

In the October accident involving Delphi, the front of its 2014 Audi SQ5 was moderately damaged when, as it waited to make a left turn, another car broadsided it, according to an accident report the company shared with AP. The car was not in self-driving mode, Delphi spokeswoman Kristen Kinley said.

Google, which has 23 Lexus SUVs, would not discuss its three accidents in detail.

The accidents are not Google's first: In a briefing with reporters a year ago, the leader of Google's self-driving car program acknowledged three others between when the company first sent cars onto public roads several years ago — without the state's official permission — and May 2014.

In a written statement, Google said that since September, cars driving on streets near its headquarters in Mountain View had "a handful of minor fender-benders, light damage, no injuries, so far caused by human error and inattention."

Google said that while safety is paramount some accidents can be expected, given that its cars have gone "the equivalent of over 15 years of typical human driving," or approximately 140,000 miles.

The national rate for reported "property-damage-only crashes" is about 0.3 per 100,000 miles driven, according to data from the National Highway Traffic Safety Administration.

In that context, Google's three in about 140,000 miles may seem high. As the company pointed out, however, perhaps 5 million minor accidents are not reported to authorities each year, so it is hard to gauge how typical Google's experience is.

Three other states have passed laws welcoming self-driving cars onto their roads. Regulators in Nevada, Michigan and Florida said they were not aware of any accidents.

As self-driving cars proliferate, others issues will arise that human drivers have dealt with for decades, notably who's liable for an accident. Each test car is required to have $5 million insurance.

Interest in accidents will remain high, especially if the self-driving car is at fault, said Bryant Walker Smith, a l

Google, Delphi Self-Driving Cars Getting Dinged in California

Updated 12:15 a.m ET Friday.

Somebody almost had to call a self-driving tow truck.

Two autonomous vehicles, one powered by Google and the other from Delphi Automotive, nearly collided this week in what’s thought to be the first close call between a pair of self-driving cars, Reuters reports.

The Google car, a modified Lexus, cut off Delphi’s modified Audi and forced it abort changing lanes to avoid a crash, Delphi’s John Absmeier told Reuters. Absmeier was a passenger in the Audi at the time of the incident; nobody was hurt in the event.

Delphi has two driverless Audi prototypes on California roads, while Google has 20 being tested.

The near collision raises important questions about self-driving cars, including who’s responsible when two are involved in an accident.

Update: A Delphi spokesperson emailed Fortune with the following statement: “During a demo drive of our automated vehicle, our expert used the interaction with the Google car as an example of the types of scenarios that the car can encounter in real-world driving. It was an anecdote of an interaction, not a ‘near miss’. Reuters completely misrepresented the facts.”

Google's self-driving Lexus cuts off self-driving Audi

There are few drivers who haven't experienced what it's like to be cut off by another motorist who nearly causes an accident. Now, two autonomous-driving cars from competing companies have matched that common human experience.

Of course, the self-driving cars weren't able to give each other the finger after it happened.

Auto parts maker Delphi says that its experimental self-driving car was on a street in Silicon Valley earlier this week and was moving into another lane when an experimental self-driving car from Google cut it off by moving into the same lane.

Delphi disclosed the incident to a Reuters reporter earlier this week and confirmed it to CNNMoney Friday. Spokeswoman Kristen Kinley insists it wasn't truly a near miss but was instead an example about how self-driving cars effectively avoid accidents.

"Our car saw the Google car move into the same lane as our car was planning to move into, but upon detecting that the lane was no longer open it decided to terminate the move and wait until it was clear again," she said.

Google agreed with Delphi's take on the incident.

"The headline here is that two self-driving cars did what they were supposed to do in a fairly ordinary everyday driving scenario," said a spokeswoman.

Related Google's new self-driving cars hit the road

Self-driving cars are allowed on streets and highways in California, but they are required to have a human driver sitting behind the steering wheel able to take over control of the car if there is a problem. The cars are not yet for sale to the public. The self-driving vehicles now on the roads are being used by companies that are developing the technology.

Self-driving cars are still rare, they're becoming more common on the streets near Google's Mountain View, Calif, headquarters. Google has 23 self-driving Lexus RX450h crossovers on the road, and recently put nine prototypes of its own self-driving car design on the road as well. Delphi, which recently drove a self-driving Audi across country, does most of its test drives near its autonomous driving lab, which is near to Google's headquarters.

Self-driving cars have been involved in a number of minor accidents with traditional driver cars during the last few years that they were being tested on the road. Google has driven the cars more than 1 million miles in autonomous driving mode. It says it has had 12 accidents. Most occurred when the autonomous driving car was stopped or nearly stopped and were struck from behind by a car with a driver.

There was a human driver sitting behind the wheel in all the cases, and in instances where the car was in motion when the accident occurred, the human driver took control of the car.

Self-driving cars from rivals Google, Delphi in close call

[Update 6/27]: In a statement from Delphi, the company has added further clarification and now claims that the two vehicles were never in any danger of colliding. The statement claims the original report from Reuters was incorrect, though the news agency has said it stands by its original report.

The most futuristic car crash in the history of humanity almost took place in Palo Alto, California, this week, when one of Google’s self-driving cars cut off another automated vehicle on a city street. Both cars were under the control of their respective computers, and they didn’t actually collide, but the close call marks a first as driverless vehicles increasingly expand beyond their limited test zones.

John Absmeier, the global business director for Delphi Automotive, was a passenger in his company’s automated vehicle prototype, a driverless Audi Q5, when it encountered one of Google’s self-driving Lexus SUVs.

Absmeier told Reuters that his vehicle was negotiating a lane change when the Google vehicle cut it off, causing the Audi to take “appropriate action” to avoid a crash.

The fact that no collision occurred is perhaps the most important takeaway from the story. Self-driving cars are appealing because of both convenience and safety. In this situation, two driverless vehicles running two different automated driving programs encountered each other in a potentially dangerous situation—one in which a human driver might have panicked—and both vehicles emerged unscathed.

Delphi’s driverless Audi

Delphi Automotive

Photo via Delphi Automotive

Near-miss between self-driving cars shows computers are smarter than us

MOUNTAIN VIEW, Calif. (Reuters) - Two self-driving prototype cars, one operated by Google Inc and the other by Delphi Automotive Plc, had a close call on a Silicon Valley street earlier this week, a Delphi executive told Reuters on Thursday.

It was believed to be the first such incident involving two vehicles specially equipped for automated driving.

The incident occurred Tuesday on San Antonio Road in Palo Alto, said John Absmeier, director of Delphi’s Silicon Valley lab and global business director for the company’s automated driving program, who was a passenger in one of the cars.

No collision took place.

Google declined to comment.

Absmeier was a passenger in a prototype Audi Q5 crossover vehicle equipped with lasers, radar, cameras and special computer software designed to enable the vehicle to drive itself, with a person at the wheel as a backup.

As the Delphi vehicle prepared to change lanes, a Google self-driving prototype - a Lexus RX400h crossover fitted with similar hardware and software - cut off the Audi, forcing it to abort the lane change, Absmeier said.

The Delphi car “took appropriate action,” according to Absmeier.

Delphi’s Silicon Valley lab is based in Mountain View, not far from Google headquarters. While Delphi is running two Audi prototypes in California, Google has been testing more than 20 Lexus prototypes.

On Thursday, Google started testing self-driving vehicle prototypes of its own design on local streets. The latest prototypes use the same software as the Lexus vehicles.

Both companies previously have reported minor collisions of self-driving cars with vehicles piloted by people. In most of those cases, the self-driving car was stopped, typically at an intersection, and was rear-ended by another vehicle.

In all cases, the self-driving prototype was not at fault, according to the California Department of Motor Vehicles and the companies.

Two rival self-driving cars have close call in California

In a showdown of self-driving cars a Google vehicle recently cut off a Delphi car attempting to switch roadway lanes, and came perilously close to crashing.

Auto parts maker Delphi says that its experimental self-driving car was on a street in Silicon Valley when the near accident occurred as both car moved in to the same lane of traffic. A Delphi spokeswoman insisted it was not so much a near-miss as it was proof that the autonomous cars are capable of avoiding accidents.

"Our car saw the Google car move into the same lane as our car was planning to move into, but upon detecting that the lane was no longer open it decided to terminate the move and wait until it was clear again," Kristen Kinley told CNN Money.

Google agreed.

"The headline here is that two self-driving cars did what they were supposed to do in a fairly ordinary everyday driving scenario," said a representative.

Self-driving cars are allowed on streets and highways in California, but they're required to have a human sitting behind the steering wheel to take charge if there's a problem. There's no indication that was the case in the near miss.

Google revealed in May that its driverless cars had been involved in 11 accidents, but they were mostly minor fender-benders and no one was injured. The company claims none of the accidents was the fault of the Google car.

Google currently has 23 self-driving Lexus cars on the road along with nine prototypes of its own car design.

Google and Delphi self-driving cars in near miss

Delphi Automotive has contradicted a senior exec at the company by denying that one of its robo-cars had come close to being pranged by a Google self-driving vehicle.

The firm told Reuters on Friday that the news wire's earlier report – which earlier this week had quoted John Absmeier, Delphi's Silicon Valley lab director, describing a close call with Google's car – was wrong.

A spokeswoman at the vendor has since claimed that "the vehicles didn't even come close to each other."

Absmeier had apparently told Reuters, which stands by its original story, that the prototype cars had been involved in a near miss.

He said, according to the report, that Google's Lexus RX400h had cut off Delphi's Audi Q5 SUV – the car Absmeier was travelling in as a passenger on San Atonia Road in Palo Alto, California at the time of the incident.

The exec added that Delphi's prototype car "took appropriate action" to avoid a collision with Google's robo-vehicle.

Delphi has now denied that a near-miss between the rival self-driving cars ever occurred. Its spokeswoman said:

During a recent visit with Reuters, our Delphi expert described an actual interaction that we encounter all the time in real-world driving situations. In this case, it was a typical lane change manoeuvre. No vehicle was cut off and the vehicles didn't even come close to each other. Both automated vehicles did exactly what they were supposed to do.

However, Delphi admitted that the Google car did force its vehicle to terminate a lane change. But the firm's spokeswoman claimed that the cars were roughly a lane width apart at the time of the incident.

Google, meanwhile, told Reuters that the "two self-driving cars did what they were supposed to do in an ordinary everyday driving scenario."

Vulture Weekend suspects that Absmeier won't be wheeled out by Delphi to speak with other journos about its tech any time soon. ®

Sponsored: Becoming a Pragmatic Security Leader

Robo-car wars: Delphi's near crash, prang, wallop with Google DENIED!

June 29, 2015

Delphi is back-pedaling after one of its executives told Reuters last week one of its self-driving vehicles had to take evasive maneuvers to avoid being side-swiped by an autonomous Google car.

Although Delphi executive John Absmeier — who was in the self-driving Audi (run by Delphi software) at the time of the incident — originally told Reuters the Delphi car “took appropriate action” to avoid being hit by a Google-operated self-driving Lexus, the company is now saying “the vehicles didn’t even come close to each other.”

“During a recent visit with Reuters, our Delphi expert described an actual interaction that we encounter all the time in real-world driving situations. In this case, it was a typical lane change maneuver. No vehicle was cut off and the vehicles didn’t even come close to each other,” the Delphi spokeswoman told Reuters. “Both automated vehicles did exactly what they were supposed to do.”

Originally, Absmeier claimed the incident took place when the Delphi vehicle he was riding in was about to make a lane change but was cut off by the Google vehicle, forcing the Audio to abandon the lane change.

Although Google did not comment after Absmeier first talked to Reuters, it later released a short statement: “two self-driving cars did what they were supposed to do in an ordinary everyday driving scenario.”

Both Google and Delphi have labs based in Mountain View. Although Delphi is testing only two Audi prototypes, Google has had more than 20 Lexus vehicles equipped with its technology on California roads.

Google last Thursday began testing its own self-driving prototypes on California streets. The Google-built vehicles use the same software and sensors as the Lexus prototypes.

Although both companies have reported accidents in the past, they were all no-fault incidents, mostly due to being rear-ended at traffic lights by another driver.

Jennifer Cowan is the Managing Editor for SiteProNews.

Delphi Changes Story, Says Google Self-Driving Car Did Not Nearly Crash Into Audi

After a report that a Delphi driverless car and a Google driverless car came close to a traffic incident on a road in Palo Alto, Calif., Delphi has issued a statement that “the vehicles didn’t even come close to each other.”

The original report, by Reuters, quoted a Delphi official as saying he was in one of his company’s driverless cars when it was “cut off” by a Google car as it was preparing to make a lane change. He said the Delphi car “took appropriate action,” aborting the lane change.

But a Delphi spokesperson subsequently said that the incident was blown out of proportion, calling it “a typical lane change maneuver” and adding that “no vehicle was cut off.” Google, which did not originally comment, put out a statement saying both cars “did what they were supposed to do in an ordinary everyday driving scenario.”

According to a spokesperson, Reuters “stands by the accuracy of its original story.”

[Reuters]

This Is What Apple’s 'Titan' Car Could Look Like Ford Ford Ford Ford Ford Ford 1 of 6 Advertisement

Contact us at editors@time.com.

Driverless Car Maker Denies Claim of Near-Miss With Google Car

It has been reported that a Delphi Automotive-owned self driving Audi was forced to take ‘appropriate action’ in order to avoid crashing into a Google self driving Lexus after it was cut off by the car.

John Absmeier was traveling in the Delphi Audi Q5 self driving car on San Antonio Road in Palo Alto, California when the clash happened. He has said that the car was forced to abort changing lanes as a result of the rival’s driving behaviour. After the incident, it was reported by Reuters as a ‘close call’ who used Absmeier’s comments in their article. But both Google and Delphi have tried very hard to say that the cars were both operating safely when the incident occurred.

Kristin Kinley, a Delphi spokeswoman, has said “The story was taken completely out of context when describing a type of complex driving scenario that can occur in the real world. Our expert provided an example of a lane change scenario that our car recently experienced which, coincidentally, was with one of the Google cars also on the road at that time”. Delphi also released a later statement saying that Reuters misrepresented the facts as the story was an “anecdote of interaction, not a ‘near miss'”.

Google has also stated that there was no threat of a collision during the incident and neither of the companies was at fault. However, Reuters have said that “the description differed from Delphi’s original account” while a spokesman has confirmed that Reuters “stands by the accuracy of its original story”. The California Department of Motor Vehicles has also backed up the stories by both companies, saying that neither was at fault during the incident.

The safety records of self driving cars are particularly protected by companies. A Google report that was published in May showed that 11 minor accidents had been suffered by the company’s fleet of self driving cars since 2009 when the program began. Chris Urmson, director of the program, said that none of the self driving cars were the cause of any incidents and many of the incidents were caused by other cars driving into the back of the Google vehicles.

Delphi has also reported one accident with its self driving cars which happened in October of last year. The report showed that the car had been broadsided by another vehicle while it was planning to make a left turn. The accident had to be reported under Californian law, but the Delphi car wasn’t actually in self driving mode at the time. By May this year, seven companies had gained self car testing permits in California, and there were 48 driverless cars on the road. None of the other five companies had reported any accidents or incidents at that point.

Google has recently announced that new purpose built self driving cars will be unleashed on public roads. The Google cars are known as ‘koala cars’ because they have a small rounded design which gives them an appearance that makes them look like the outback marsupial animal. In a Google blog post, the company explained more of the features of the cars- “They’re ultimately designed to work without a steering wheel or pedals, but during this phase of our project we’ll have safety drivers aboard with a removable steering wheel, accelerator pedal, and brake pedal that allow them to take over driving if needed. The prototypes’ speed is capped at a neighbourhood-friendly 25mph, and they’ll drive using the same software that our existing Lexus vehicles use”.

Self driving cars use a combination of lasers, radar, cameras and software to enable them to ‘see’ the road around them and move about safely. The self driving cars are limited to a slow speed which prevents them from being able to drive autonomously on motorways. While Google has a fleet of 20 Lexus cars, Delphi has been testing just two prototypes in California. Self driving cars are predicted to be the future of driving, and so the current prototype testing may lead the way for models that we will all be riding in the future.

Google and Delphi Self Driving Cars Clash In Near Miss Accident

One of the main reasons behind Google’s commitment to driverless cars is their ability to avoid accidents and improve traffic safety , as they eliminate human error, which is the single most common cause of motor vehicle crashes. Up until a few weeks ago, the tech giant’s self-driving car prototypes had covered a total of 1.8 million miles without causing any accident, but their crash-free record was on a brink of being tarnished recently, with a report claiming that one of the cars almost hit another vehicle while being operated on a public road in California.

As reported by Reuters, one of Google’s driverless cars nearly collided with another car while driving in Silicon Valley, but an accident was avoided thanks to a timely maneuver performed by the tech company’s car. The fact that the other vehicle involved in this close encounter was also a self-driving car, makes the situation that much more intriguing. The other vehicle in question was an Audi Q5 crossover fitted with self-driving equipment developed by Delphi Automotive. Delphi’s vehicle was driving along a street in Palo Alto, when it started performing a maneuver to change lanes, but was cut off by Google’s car – a Lexus RX400h, and was forced to return to its lane and try to perform the lane change later.

Even though an accident was avoided, this close call got a lot of media attention, as it reveals that there is still a certain risk of autonomous vehicles getting involved in potentially dangerous traffic scenarios. On the other hand, it also shows that driverless cars are capable of detecting traffic situations that could jeopardize the safety of their occupants and other road users, and taking appropriate action to avoid them.

Delphi was quick to deny the reports of a near collision and explain that the reported incident actually reaffirms the ability of self-driving cars to avoid accidents. John Absmeier, director of Delphi’s Silicon Valley lab, was a passenger in the Audi Q5 when the incident took place, and later said that it wasn’t nearly as serious as media tried to portray it.

“The story was taken completely out of context when describing a type of complex driving scenario that can occur in the real world. Our expert provided an example of a lane change scenario that our car recently experienced which, coincidentally, was with one of the Google cars also on the road at that time. It wasn’t a ‘near miss’ as described in the Reuters story,” said Delphi in a statement.

In any case, no matter how harmless this incident was, it is bound to raise more concerns over whether autonomous cars are really able to operate on public roads without causing any accidents and without being a threat to public safety, and it might well have a negative impact on the public perception of self-driving cars.

Google's Driverless Car Almost Hit Delphi's Autonomous Vehicle, Report Says

Similar Incidents

By textual similarity

Did our AI mess up? Flag the unrelated incidents