Autonomous Vehicles, LEAF and others...

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Via IEVS:
Tesla Model 3 Outperforms In New IIHS Level 2 Autonomy Tests
https://insideevs.com/tesla-model-3-outperforms-in-new-iihs-level-2-autonomy-tests/

However, it’s not perfect.
With drive assistance features becoming standard in many vehicles, it’s good to remember they’re not all the same. Some work better than others, and the Insurance Institute for Highway Safety is devising a series of tests to evaluate these systems in various driving situations. The focus of the tests is Level 2 “autonomy” as defined by SAE International, which includes adaptive cruise control (ACC) and active lane-keeping (ALK).

In IIHS’s research, the company tested five vehicles: 2017 BMW 5 Series with “Driving Assistant Plus,” 2017 Mercedes-Benz E-Class with “Drive Pilot,” 2018 Tesla Model 3 and 2016 Model S with “Autopilot” (software versions 8.1 and 7.1, respectively), and 2018 Volvo S90 with “Pilot Assist.” The Tesla Model 3 performed best through the bevy of tests – however, it wasn’t perfect.

The cars were put through four different scenarios to test their ACC. The first involved driving at 31 miles per hour toward a stationary vehicle with ACC off and autobrake turned on. Only the two Teslas failed, hitting the stationary target. The same test was then performed with ACC on, and the Tesla Model 3 slowed with gradual decelerations. All vehicles passed this portion.

A third scenario had the cars follow a lead vehicle that slowed to a stop and then accelerated. Every car performed well in this test. The final test had the test cars follow a lead vehicle that changed lanes to reveal a stationary vehicle in the test vehicle’s path. The vehicles had about 4.3 seconds before colliding with the stationary vehicle. However, all the test cars performed well with none of the vehicles striking the stationary vehicle.
One wonders if recent updates have allowed the Teslas (as well as the others) to pass this last test, as we know Tesla Model S/Xs have hit stopped vehicles in just this situation. Good news if this is something repeatable in the real world.

Where the Tesla Model 3 truly outshined its competition was in the hill and curve tests for ALK. Here, the IIHS conducts six tests on three different sections of curved roads. Only the Model 3 stayed within its lane through all 18 trials. To test how the ALK of all five vehicles performed on hills, the IIHS mapped out a course on three hills with different slopes, running six different tests on each hill in each vehicle. Here, the Model 3 had just one deficiency, touching the centerline once in 18 tests.

The Model 3’s competitors had various levels of success with the ALK tests. The BMW, Mercedes, Model S, and Volvo all went over the centerline during both the hill and curve test, with the Model S crossing the most at 12 times when being tested on the hills. The Volvo crossed the line the most at eight times during the curve test. Some of the vehicles, such as the 5 Series, E-Class and S90 had the ALK system disengage during the tests, with the 5 Series system disengaging the most. . . .

“We’re not ready to say yet which company has the safest implementation of Level 2 driver assistance, but it’s important to note that none of these vehicles is capable of driving safely on its own,” David Zuby, IIHS chief research officer, says in the study. “A production autonomous vehicle that can go anywhere, anytime isn’t available at your local car dealer and won’t be for quite some time. We aren’t there yet.”
 
Via IEVS:
IIHS: Fatal Tesla Crash Proves Partial Automation Is Risky
https://insideevs.com/iihs-tesla-crash-partial-automation-risky/

A headline which should elicit a "Well, Duh!" from anyone who's familiar with the past half-century or so of research in the area.

Some quotes:
David Aylor has a job with some pretty interesting perks. Recently, the Insurance Institute for Highway Safety (IIHS) manager of active safety testing has been putting miles on a Tesla Model S in an attempt to gauge the effectiveness of its Autopilot advanced driver-assistance system (ADAS). He’s found that the system has some flaws, though, if you’ve been following coverage of a handful of high-profile crashes involving the Tesla vehicles being operated on Autopilot, you likely already knew that.

In particular, Aylor points to the system’s occasional failure to properly handle road splits. It seems that there can be some confusion as to which lines to follow, and if drivers aren’t paying attention, it can lead to a collision. This appears to have been the case in one of the most famous of incidents: the crash that claimed the life of Walter Huang. That incident is outlined in a recent report about autonomous vehicles. . . .

Aylor also brings up a similar incident that was filmed by a driver in Chicago testing for this very situation not long after the Huang crash. In that case, the video of which we’ve embedded below, the car doesn’t seem to know which set of lines to follow, and the driver has to intervene, braking just in front of the gore point.

These incidents, IIHS says, are evidence of the risk that partial autonomous systems can pose. Despite the fact that they found Tesla Autopilot can reduce injuries and damage claims, it is also true that it’s not a perfect system and drivers need to be alert and ready to take control if it runs into trouble. And the problem isn’t limited to Tesla.

The report notes that vehicles from other automakers equipped with Level 2 ADAS systems have also been involved in crashes. Those occurrences, however, haven’t made headlines like those involving the Silicon Valley company. For whatever reason, none of those incidents made it into this particular report either

The report doesn’t offer much in the way of analysis, but the reason for the danger seems clear. Drivers, used to a system that works perfectly a very high percentage of the time, can be caught off guard when suddenly it experiences difficulty. . . .
Direct link to IIHS report: http://www.iihs.org/iihs/sr/statusreport/article/53/4/2
 
GRA said:
Via IEVS:
A headline which should elicit a "Well, Duh!" from anyone who's familiar with the past half-century or so of research in the area.

The headline was probably trying to attract a larger audience than this select group, but yeah, I agree. It still looks like 'self-driving' software isn't quite ready for prime-time. Kind of a bummer for me since I'm hoping to be driven around in my dotage, which is no longer that far off...
 
Via GCC:
Autonomous electric shuttle in long-term public demo in Candiac, Québec
http://www.greencarcongress.com/2018/08/20180813-candiac.html

The City of Candiac, Québec is hosting a long-term demonstration project for a 100% electric autonomous shuttle to be used on public roads for the first time in Canada. . . .

With a capacity of 15 passengers and operating at a maximum speed of 25 km/h (15.5 mph), the NAVYA autonomous electric shuttle facilitates travel and promotes the use of public transit in underserved areas.

Starting in September, the autonomous electric shuttle will provide citizens with free transportation between the park-and-ride lot and André-J.Côté Park, making its way along Montcalm Boulevard North. Stops include: City Hall, the Chartwell Le Montcalm senior's residence and several businesses.

An operator will be on board the shuttle for the duration of the project to answer any questions that users might have. This operator will also be able to take control of the shuttle as required. The shuttle will run throughout the fall until winter weather conditions take over. An experimentation phase, with no passengers, will then begin to ensure the shuttle’s performance during the winter season. . . .

The manufacturer NAVYA designs, produces and markets autonomous, driverless, electric vehicles that combine robotic, digital and automotive technologies. NAVYA has a range of two autonomous vehicles: the AUTONOM SHUTTLE, launched in September 2015, of which more than 100 have already been produced as of today and 89 sold in 17 countries and the AUTONOM CAB, unveiled in November 2017 and whose first road tests will start shortly.
 
Via ABG:
Kroger begins tests of driverless grocery delivery in Arizona
Would you pay $5.95 for bread from an autonomous breadbox?
https://www.autoblog.com/2018/08/16/kroger-tests-driverless-grocery-delivery-arizona/

U.S. supermarket giant Kroger said it will start testing driverless grocery delivery on Thursday with technology partner Nuro at a single Fry's Food Store in Scottsdale, Arizona.

Kroger and rival Walmart each have teamed up with autonomous vehicle companies in a bid to lower the high-cost of "last-mile" deliveries to customer doorsteps, as online retailer Amazon.com rolls out free Whole Foods delivery for subscribers to its Prime perks program. . . .

The first phase of the test will use a fleet of Toyota Prius cars equipped with Nuro technology. Those cars have seats for humans who can override autonomous systems in the event of an error or emergency. Nuro's R1 driverless delivery van, which has no seats, will begin testing this autumn, the companies said.

"While we compete final certification and testing of the R1, the Prius will be delivering groceries and helping us improve the overall service," a Nuro spokeswoman said.

Self-driving car delivery from the Fry's store will cost $5.95 with no minimum order. It is only available at addresses within the store's ZIP code of 85257, Kroger said. . . .
 
Via GCC:
Hyundai Motor completes South Korea’s first domestic autonomous truck highway journey
http://www.greencarcongress.com/2018/08/20180824-hyundai.html

. . . Hyundai’s Xcient truck drove approximately 40 km (25 miles) on the highway between Uiwang and Incheon, carrying a large semi-trailer simulating cargo transportation.

  • This successful demonstration proves that innovative autonomous driving technology can be used to transform the trade logistics industry. At this stage, a human driver is still used to control the vehicle manually in certain situations, but I think we will achieve level 4 automation soon as we are constantly upgrading our technological capability.

    —Maik Ziegler, Ph.D., Director of Commercial Vehicle R&D Strategy Group at Hyundai Motor Company

The demonstration, which took place on 21 August, was conducted using Hyundai’s Xcient model truck, which has a maximum load capacity of 40 tons. This was semi-equipped with a Society of Automotive Engineers (SAE) standard Level 3 autonomous driving system, enabling it to steer, accelerate or decelerate, and maneuver through traffic, all without human input. A human driver was on-board to take over manual control when required.

The vehicle was equipped with innovative technology features, which enabled it to maintain and change lanes during the natural flow of traffic, detect lane changes made by vehicles in front of it, navigate through tunnels, and perform a complete halt or accelerate according to road traffic. . . .

The test route is Hyundai Glovis’s parts transportation most frequently travelled section for vehicles heading to the Port of Incheon. This includes 40km in total of automobile highway. The truck successfully completed the journey, travelling 40km in 1 hour, while abiding strictly to the expressway speed limit of 90 km/h (56 mph).

Hyundai Motor is planning to undertake further autonomous navigation technology tests in future in a variety of areas like Busan, and plans to concentrate on its enterprise development capabilities with the aim of early commercialization of the technology.

Expressways headed toward the Port of Incheon display heavy traffic even during weekdays, due to a high quantity of goods being exported. Therefore the vehicle’s autonomous technology and know-how had to be sufficient enough to adapt to unprecedented situations throughout the journey. . . .

10 different sensors, including 3 front and side-rear cameras, 2 frontal and rear radars, 3 Lidars in the front and sides, and a hitch angle sensor in the trailer coupler which computes the change in angle between the truck and trailer in real-time, allowing the truck to be safely stabilized upon sharp turns.

The data collected by each sensor collaborates with the HD map and sends information to the electronic control module for localization. The module makes accurate decisions for each situation, controlling the speed, steering, and breaking accordingly. . . .
 
https://www.theverge.com/2018/8/21/17762326/waymo-self-driving-ride-hail-fleet-management was interesting. It linked to some interesting stuff like the below (not sure if it's been pointed to already in this thread):
https://www.theverge.com/2018/5/9/17307156/google-waymo-driverless-cars-deep-learning-neural-net-interview had a pointer to a semi-funny Waymo-related video at https://twitter.com/nitguptaa/status/990683818825736192
https://www.theverge.com/2017/10/31/16579180/waymo-self-driving-test-facility-castle-google

One of the articles points to https://www.theinformation.com/articles/inside-cruises-bumpy-ride-the-limits-of-self-driving-cars but unfortunately, it's sorta behind a paywall. You can enter an email address to read it. The article has a lot of interesting corner cases and limitations supposedly of Cruise Automation's efforts as of March 2018 that I wasn't aware of.

It also had a sobering section near the end:
‘90-90’ Rule

Cruise, like many other developers, is trying to follow Google. Google began its self-driving car project in 2009, and by 2010 had successfully driven autonomously through 10 separate 100-mile loops in California, across urban streets and highways. In early 2017, Google spun out the venture as Waymo, a separate unit of Google owner Alphabet, after making slow progress in commercializing the technology.

Waymo has an explanation for that: “When you’re 90% done you still have 90% to go,” said Sacha Arnoud, a director of engineering at Waymo, in a recent talk at the Massachusetts Institute of Technology. In other words, writing the first 90% of the code that’s necessary for self-driving cars to function in the real world takes just 10% of the time. Completing the job, or the last 10%, requires 10 times the initial effort.

“You need to 10x the capabilities of your technology. You need to 10x your team size, including finding effective ways for more engineers and more researchers to collaborate together. You need to 10x the capabilities of your sensors. You need to 10x the fundamentally the overall quality of the system, and your testing practices,” he said.
Also, not sure if these have been posted before:
https://www.cnbc.com/2018/05/14/alphabet-waymo-tells-police-how-to-break-in-disable-self-driving-cars.html
https://www.recode.net/2017/10/15/16472896/alphabet-waymo-self-driving-law-enforcement-crashes-accidents

These obviously will have implications when self-driving cars are driving around w/no occupants.
 
'I hate them': (Phoenix area) Locals reportedly frustrated with Alphabet's self-driving cars
https://www.cnbc.com/2018/08/28/locals-reportedly-frustrated-with-alphabets-waymo-self-driving-cars.html
 
I've been seeing both Waymo and Uber autonomous vehicles in my city recently, several times each in the past couple of weeks. I don't know if they're just being used to generate high-def LIDAR maps for now, or if they're actually being driven autonomously; they do have humans behind the wheel. As this is outside of Silicon Valley, I guess they're getting ready to expand outwards.
 
My city has just added a security robot to one of the parking garages downtown. Looks the same as this: https://www.knightscope.com/knightscope-k5/

There's a portable inductive charging pad that it sits on when not making its rounds; I'm considering moving the pad the next time I see it in motion to see if it can find it on its own. I've been seeing it immobile for a couple of days, but last night walking back from the grocery store I finally saw it moving around, followed by a crowd of teens/20-somethings who were taking selfies with/videos of it. It's a multi-story garage, and I doubt it can make it up the ramp to the 2nd or 3rd floors, but it circuited the ground floor okay, kind of like a slower Roomba without the bumping into things. I walked in front of it a couple of times and it detected me okay and stopped. I think it looks pretty top heavy, and can easily see the next urban version of cow-tipping involving these (now that tipping over Smart cars seems to have died down); balaclavas/face masks required.

Added: Last night I saw it start up the ramp to the second floor, but it stopped about 1/3rd of the way up, possibly due to lack of battery power, but IMO more likely due to the large transverse joint in the concrete that spans full width of the ramp, and which might cause it to get stuck or else topple.
 
Via GCC:
Bosch IAA CV survey: Germans would increasingly feel safer with autonomous trucks on the road
http://www.greencarcongress.com/2018/09/20180914-bosch.html

. . . While almost 40% of respondents would rather that trucks have a human driver at the wheel, already more than one in three (37%) no longer have a preference for a human over a machine. One in four respondents would have more confidence in an autonomous truck than in a human driver.

For now, driverless trucks are still an unrealized vision. But the survey shows that in Germany, people increasingly favor automated trucks when it comes to safety. The intelligent technology on board such trucks could prevent a large number of accidents; nine out of ten accidents are due to human error. . . .

At present, most people stuck in traffic find trucks and vans rather annoying. According to 57% of respondents, Germans feel particularly unsafe in critical situations involving trucks—for instance, when merging onto the freeway or when a truck is turning.

More than one in two (56%) believe that there are too many road freight vehicles on the road. Around half of respondents said their biggest complaint is when trucks block traffic while parking. Other annoyances include commercial-vehicle emissions (50%) and truck noise (43%). Only one in five respondents said that truck traffic didn’t bother them.

What the survey also highlights is that very few people are willing to do anything themselves to relieve delivery traffic on the road. Three-fourths of Germans (73%) don’t want to shop less online. Few of them (49%) are willing to compromise by accepting longer waiting times for parcel deliveries as a way to relieve traffic—having parcel delivery just once a week instead of every day.

However, one in four respondents (27%) did say that they would reduce delivery traffic by returning fewer goods, while 36% would have their parcels delivered to a central parcel station or collection point and then pick them up themselves.

Paying more for parcels to be delivered—to have, say, more evening deliveries so as to spread traffic throughout the day—is something only 15% of respondents would consider.
Of course, the first time such a vehicle causes a serious or fatal crash the public will likely lose a lot of confidence in them, just as has happened here (e.g. Uber crash in Arizona), so slow and cautious introduction is the way to go.

Also GCC:
Volvo Trucks presents Vera autonomous electric vehicle for future transport solutions
http://www.greencarcongress.com/2018/09/20180913-vera.html

. . .Vera is controlled and monitored via a cloud-based service, and has the potential to make transportation safer, cleaner and more efficient. The driveline and battery pack are of the same type that are used in Volvo Trucks’ electric trucks.

Volvo’s long-term goal is to offer companies that need continuous transport services between fixed hubs a complement to today’s offerings.

Growing world population and increasing urbanization are leading to significant challenges to solve environmental issues such as congestion, pollution and noise. Rising consumption, the fast growth of e-commerce and the wide-spread shortage of drivers put higher demands on efficient transport solutions. . . .

Volvo Trucks’ Vera future transport solution is intended to be used for regular and repetitive tasks characterized by relatively short distances, large volumes of goods and high delivery precision. Transports between logistic hubs are typical examples, but additional use cases could also be applicable. . . .
 
Via GCC:
Daimler and Bosch jointly premiere Automated Valet Parking in China
http://www.greencarcongress.com/2018/09/201800915-daimlerbosch.html

Daimler and Bosch announced the successful premiere of their joint Automated Valet Parking pilot in Beijing. The infrastructure-supported driverless parking technology, which made its debut in the parking garage at the Mercedes-Benz Museum in Stuttgart in 2017 (earlier post), marks the first pilot of its kind in China.

Automated Valet Parking, to the extent legally allowed where used, enables vehicles to proceed to an assigned space and for the user to retrieve the vehicle via their smartphone. This will undergo further testing by Daimler and Bosch at the Mercedes-Benz Research and Development Center. . . .

Automated Valet Parking begins when the user parks the vehicle in a designated drop-off area before sending it to be parked using a smartphone app. After being registered by the intelligent system infrastructure installed in the car park, the vehicle is started and guided to an assigned parking space.

Sensors installed in the car park monitor the driving corridor and its surroundings while steering the vehicle. The vehicle’s onboard technology safely maneuvers it in response to commands from the intelligent car park infrastructure, stopping the vehicle in good time when necessary. When a user is ready to pick up the car, it can be called through a smartphone app, after which it rolls to the pick-up area without a driver.

The pilot in Beijing demonstrated Automated Valet Parking’s upgraded and more practical, real-world functionality. At the event, two vehicles were tested simultaneously to mirror the unique and complex traffic conditions found in contemporary Chinese cities.

Both vehicles were also able to successfully navigate to a service area that could be equipped with a diverse range of facilities in the future. These might include charging infrastructure, car washing stations, express package pick-up and other features designed to meet the unique needs of Chinese customers.

Automated Valet Parking is an efficient solution for both car park management and vehicle users. Car parks equipped with this intelligent infrastructure as far as legally allowed can potentially accommodate up to 20% more vehicles, while users save time and enjoy greater convenience. . . .
While I can see humans doing the charging in areas with low wages, in the developed world it seems almost guaranteed that charging will be automated, either through inductive charging or else the robot conductive charging that Tesla and others have been working on. I have little faith in the latter method becoming both reliable and cheap enough to commercialize any time soon.
 
Via ABG:
VW trying to get automakers to agree on self-driving car standards
Meeting an industry standard could reduce liability for accidents
https://www.autoblog.com/2018/09/14/vw-automakers-self-driving-car-standards/

. . . The Volkswagen Group is discussing the initiative with more than 15 potential partners, according to the report, which cited an anonymous company executive. It did not identify any of the other companies involved.

Carmakers, suppliers and technology companies are already investing more than 55 billion euros ($64 billion) globally in autonomous driving, consulting firm AlixPartners estimates.

Manufacturers including BMW have indicated they would welcome common standards on aspects of self-driving cars including the configuration of their sensors.

By pooling such blueprints, they could hope to reduce costs as well as their individual accident liabilities in relation to shared technological choices.

"When you are involved in an accident, you have a better chance in court when you can prove that your car adheres to the latest technical standard," an industry executive told Automotive News. . . .
 
Via ABG:
Ford calls for industry standard signal for self-driving vehicle intent
It envisions a universal signal to communicate what an AV is doing
https://www.autoblog.com/2018/10/02/ford-industry-standard-signal-self-driving-vehicle/

Ford on Tuesday put out the call to developers of self-driving vehicles to help build a new industry standard interface to help communicate the intent of autonomous vehicles to passengers, pedestrians and other drivers sharing the road.

In a blog post published on Medium, Ford said it's made its own work in the area available to other teams through a memorandum of understanding, meaning essentially that it wants to open-source the development of a new universal AV communications interface.

"We want everyone to trust self-driving vehicles — no matter if they are riders in these vehicles themselves or pedestrians, cyclists, scooter users or other drivers sharing the road," John Shutko, Ford human factors technical specialist for self-driving vehicles, wrote in the post. "Having one, universal communication interface people across geographies and age groups can understand is critical for the successful deployment of self-driving technology."

The move builds on real-world research Ford undertook with Virginia Tech Transportation Institute starting last year in which engineers mounted a light-bar to the top of the windshield of a Ford Transit Connect van, with drivers essentially disguised as seats to simulate the van operating on its own to see whether the signals were effective. It used three different lighting scenarios: two white lights moving side to side to designate the vehicle was yielding and about to stop; a solid white light to indicate the vehicle was proceeding on its current course; and a rapidly blinking white light to communicate that the vehicle was about to accelerate from a stop. Ford outfitted the Transit Connect with multiple cameras to record video from every direction to gauge people's reactions over the course of more than 2,000 miles of testing.

It also conducted a study using virtual reality in which participants were placed at a street corner to observe a complex mix of vehicles passing through an intersection, some with the intent-interface light signals and some without.

Ford says it plans to take what it's learned and equip the self-driving intent interface on a fleet of autonomous Fusion Hybrid development vehicles to be used by Argo AI in and around Miami for more testing. Research is also under way in Europe "so we can ensure they are universally understood across regions and cultures."

The automaker is also working with the International Organization of Standardization, or ISO, and the Society of Automotive Engineers to create a unified communication interface for AVs. The goal is to reach agreement over where to place the signals on a self-driving vehicle, how to design the signals, and what colors the signals should be.
There are videos. This is going to be an issue with AVs, trying to replace a lot of the non-verbal communication that goes on between human drivers, pedestrians, cyclists etc., from eye contact through small or large physical motions (nods,head shakes, waves and so on).
 
Honda commits $2.75 billion to build autonomous vehicles with GM’s Cruise
https://techcrunch.com/2018/10/03/honda-commits-2-75-billion-in-partnership-with-gms-cruise/

Along the lines of what Ford is proposing, when Nissan showed the IDS Concept in late 2015, it also included a display and lights to show intent/give others around it cues. Skip to ~2:25 of https://www.youtube.com/watch?v=h-TLo86K7Ck.
 
Via CR:
Cadillac Tops Tesla in Consumer Reports' First Ranking of Automated Driving Systems

CR finds that these features make driving easier but introduce new safety risks
https://news.consumerreports.org/au...llac-tops-tesla-in-automated-systems-ranking/

CR rated Supercruise better because:
In Consumer Reports’ first-ever ranking of partially automated driving systems, Cadillac’s Super Cruise (shown above) was top-rated because our testing shows it does the best job of balancing high-tech capabilities with ensuring that the car is operated safely and that the driver is paying attention.

CR experts stress that the systems are not intended to be self-driving features. However, in the right circumstances, such as on long highway drives or in stop-and-go traffic, they can help relieve driver fatigue and stress.

The risks come if automakers allow the systems to operate in situations where they can’t do so safely and if the systems make it easy for drivers to feel like they don’t need to pay attention.

In CR’s rankings, Tesla’s Autopilot came in second, followed by Nissan/Infiniti’s ProPilot Assist and then Volvo’s Pilot Assist system.

Autopilot scored highly for its capabilities and ease of use, while Nissan’s system was better at keeping drivers engaged. Volvo scored comparatively lower.

IEVS, reporting on the above article:
. . . In a comparison test of a number of systems by Consumer Reports, Cadillac’s Super Cruise actually came out on top. And this isn’t the first comparison of these two systems have reached that same conclusion. So, what is going on here? Let’s take a closer look.

In everyday usage, there’s no doubt that many would prefer Autopilot, especially if they are Tesla owners familiar with its proficiency at reading the roads and adjusting to traffic accordingly. And, despite it being referred to as a beta system and recommended for highway use only, it can be engaged anywhere*. Did we mention that the system is constantly being improved upon and given even more functionality over time?

Super Cruise, on the other hand, actually scores lower than Autopilot when it comes to basic functionality. It’s just not as capable and it’s also geo-fenced, meaning that it can only be used on certain highways*. When it first rolled out, the number of roads where it would even turn on as quite limited, but that’s has grown over time and it now functions on a wide swath of our highway system.

So how does the Cadillac system manage to beat out Tesla’s? It all comes down to safety and it seems here, Autopilot is a victim of its own success. Experts in the field contend that with increased confidence in a system, comes a greater liklihood a driver will be less attentive. Sure, the Tesla system does nag you visually and audibly if you stay hands-off for 20 seconds or so, but a car can cover a lot of ground in that time* and, perhaps, find itself in a situation where it needs the driver to intervene.

By contrast, the Super Cruise system employs a camera trained on the driver and monitors their eyes to see if they look away from the road. Close your eyes or turn your head for more than four seconds and the Caddy’s steering wheel lights up and it gives an audible warning. The seat may even vibrate. . . .
https://insideevs.com/autopilot-super-cruise-compared-consumer-reports/


*The asterisked quotes are primarily why Joshua Brown died, because although A/P (and other systems using Mobileye's hard/software) had no capability to recognize crossing traffic as a real target and trigger AEB (a fact that only came out as a result of the investigation, and wasn't known to the public prior to that), Tesla just recommended in the owner's manual that A/P only be used on limited-access highways (with no at-grade crossings), but didn't prohibit it being used anywhere else, as they could have and Cadillac does. Result, one dead Tesla owner. Even with currently shortened A/P hands-off warning times, an A/P Tesla could still collide with that crossing semi, because it was only in sight for a maximum of 11 seconds from Brown cresting the hill, so if the driver had been hands-on immediately prior to that and then took their eyes off the road, there'd still be no warning given before impact.

IDK if any improvements have been made in the TACC/AEB systems now such that a similar crossing target would be correctly recognized and AEB engaged to stop the vehicle, but to my knowledge no company has made such a claim yet, and I applaud Cadillac for prohibiting Supercruise use in that situation until it does. AFAIC, NHTSA should mandate such prohibitions on all these systems, until such time as each one demonstrates that it can routinely handle crossing traffic in the straight crossing path and left turn across path situations. There's no excuse for companies using customers as Beta test guinea pigs in cases where failure = severe injury/death.
 
cwerdna said:
Waymo autonomously drives one million miles in a month, 25K miles per day
https://9to5google.com/2018/07/20/waymo-one-million-miles-self-driving/

Waymo’s autonomous cars have driven 8 million miles on public roads
https://www.theverge.com/2018/7/20/17595968/waymo-self-driving-cars-8-million-miles-testing
Waymo's now hit 10 million miles on public roads:
https://arstechnica.com/cars/2018/10/waymo-has-driven-10-million-miles-on-public-roads-thats-a-big-deal/
https://www.theverge.com/2018/10/10/17958276/waymo-self-driving-cars-10-million-miles-challenges
https://techcrunch.com/2018/10/10/waymos-self-driving-cars-hit-10-million-miles/
 
https://youtu.be/4dlkj-gzOsU was published recently. Not a lot of detail though but not surprising given that it’s mainstream news media.
 
Drivers Wildly Overestimate What 'Semiautonomous' Cars Can Do
https://www.wired.com/story/semi-autonomous-systems-safety-research-euro-ncap-thatcham/
 
Back
Top