Blue roads and glowing signs—how this startup’s tech lets cars see the world
I learned some things from reading this...
I learned some things from reading this...
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Car Progress
- Thread starter stopcrazypp
- Start date
-
- Tags
- Autonomous Vehicles
Ghost in Musk's machines: Software bugs' autonomous joy ride
Suggestion that autonomous software needs to be more rigorous and robust. Am not a developer, but would think there is already a class of software that has these attributes - like what's used for airplane control systems. Also article has some interesting observations about challenges with neural network development.
Suggestion that autonomous software needs to be more rigorous and robust. Am not a developer, but would think there is already a class of software that has these attributes - like what's used for airplane control systems. Also article has some interesting observations about challenges with neural network development.
One of the companies in the article is working with TomTom. Tesla recently switched to TomTom for their maps. Just a point I noticed, although it may not mean much.
Ultimately it's going to come down to what an accept loss is.
Do we make perfect the enemy of good enough? Especially at a time when humans are becoming worse drivers? Where we're distracted by the very type of technologies that are making autonomous cars a possibility?
What do we define as a goal to start from? The starting point is extremely important because what makes autonomous driving so attractive is you only have to solve a problem ONCE. It's not like humans where you constantly have to teach the same lessons over, and over.
Once autonomous cars have a starting point to work from you'll see the safety rapidly improving. At least until it reaches some plateau where you can't get any safer without huge expense.
Public acceptance is likely the biggest obstacle of self-driving cars. But, it's a bit of a catch-22 because the public tends to have a much larger acceptance of things that convenience them versus something they view as a novelty.
To really add convenience means the infrastructure needs to get better to capitalize on self-driving cars. Simply putting autonomous cars on the road doesn't really give us a whole lot. Sure it will improve safety, but not efficiency. To really gain efficiency means you have to capitalize on what self-driving cars bring to the table. They bring to the table the ability to constantly follow rules thereby preventing situations that harm traffic flow. They also can follow each other where they create a mini train of sorts. Thereby increasing the efficiency of the roadway.
Last edited:
Very nice.
A little surprised to see Level 3 being referred to as a car to sleep in, though. But still, a valuable effort and provides also some insight into what Tesla may be doing with their own system - and what belief-system is probably driving Elon as well.
We shall see if they can crack it that way. I'm certainly always rooting for the very small guy, like comma.ai...
Back to Baidu (the Chinese search giant), I don't see anything in solid in terms of current progress but while doing some digging I found:
Apollo
Q. Why is Baidu sharing its secret self-driving sauce? A. To help China corner the market
Baidu Releases Apollo 1.5 (New Open-Source Autonomous Driving Platform) & Announces $1.5 Billion Fund
Baidu Leads China’s Self-Driving Charge
China’s Baidu opens self-driving lab in Silicon Valley
Self-driving cars now roam across former California military facilities - mentions Baidu, Waymo, Honda, Uber and others
Autonomous Driving Unit points to a press release about a test milestone in Dec 2015 at Baidu Autonomous Car Reaches Milestone in Beijing.
Apollo
Q. Why is Baidu sharing its secret self-driving sauce? A. To help China corner the market
Baidu Releases Apollo 1.5 (New Open-Source Autonomous Driving Platform) & Announces $1.5 Billion Fund
Baidu Leads China’s Self-Driving Charge
China’s Baidu opens self-driving lab in Silicon Valley
Self-driving cars now roam across former California military facilities - mentions Baidu, Waymo, Honda, Uber and others
Autonomous Driving Unit points to a press release about a test milestone in Dec 2015 at Baidu Autonomous Car Reaches Milestone in Beijing.
Last edited:
https://storage.googleapis.com/sdc-prod/v1/safety-report/waymo-safety-report-2017-10.pdf
Waymo Safety Report 10/2017
Sensor Range
Our vehicles can see up to 300 meters away (nearly three football fields) in every direction.
Backup Computing
A secondary computer in the vehicle is always running in the background and is designed to bring the vehicle to a safe stop if it detects a failure of the primary system.
Backup Collision Detection and Avoidance System
Multiple backup systems—including independent collision avoidance systems—constantly scan the road immediately ahead and behind the vehicle for objects such as pedestrians, cyclists, and other vehicles. These redundant systems slow or stop the vehicle in the rare event that the primary system does not detect or respond to objects in the path of the vehicle.
Backup Steering
The steering system features a redundant drive motor system with independent controllers and separate power supplies. Either one can manage steering in the case that a failure occurs in the other.
Backup Braking
If the primary braking system fails, we have a full secondary braking system that immediately kicks in. Either braking system can bring the vehicle to a safe stop if a failure occurs in the other.
Backup Power Systems
Independent power sources are provided for each of the critical driving systems. These independent power sources ensure that our vehicles’ critical driving components remain online during single power failures or circuit interruptions.
Redundant Inertial Measurement Systems for Vehicle Positioning
Redundant inertial measurement systems help the vehicle accurately track its motion along the road. These two systems cross-check each other and assume control from one another, if a fault is detected in either system.
Waymo Operations Center
Waymo’s system can detect when it has been involved in a collision and will notify our Waymo operations center automatically. There our trained specialists can initiate post-crash procedures, which include procedures for interacting with law enforcement and first responders, and for sending members of our team on location. Our operations center also has rider support specialists, who can communicate directly with our passengers through our in-vehicle audio system.
Following a collision, we’re able to analyze all available data, including video and other sensor data, to evaluate factors that may have contributed to the incident, and we’re able to make any appropriate software changes and update every vehicle in our fleet accordingly. Any damage our vehicles sustain in a collision is repaired and the vehicles are tested for safety before they return to the road.
Waymo Safety Report 10/2017
Sensor Range
Our vehicles can see up to 300 meters away (nearly three football fields) in every direction.
Backup Computing
A secondary computer in the vehicle is always running in the background and is designed to bring the vehicle to a safe stop if it detects a failure of the primary system.
Backup Collision Detection and Avoidance System
Multiple backup systems—including independent collision avoidance systems—constantly scan the road immediately ahead and behind the vehicle for objects such as pedestrians, cyclists, and other vehicles. These redundant systems slow or stop the vehicle in the rare event that the primary system does not detect or respond to objects in the path of the vehicle.
Backup Steering
The steering system features a redundant drive motor system with independent controllers and separate power supplies. Either one can manage steering in the case that a failure occurs in the other.
Backup Braking
If the primary braking system fails, we have a full secondary braking system that immediately kicks in. Either braking system can bring the vehicle to a safe stop if a failure occurs in the other.
Backup Power Systems
Independent power sources are provided for each of the critical driving systems. These independent power sources ensure that our vehicles’ critical driving components remain online during single power failures or circuit interruptions.
Redundant Inertial Measurement Systems for Vehicle Positioning
Redundant inertial measurement systems help the vehicle accurately track its motion along the road. These two systems cross-check each other and assume control from one another, if a fault is detected in either system.
Waymo Operations Center
Waymo’s system can detect when it has been involved in a collision and will notify our Waymo operations center automatically. There our trained specialists can initiate post-crash procedures, which include procedures for interacting with law enforcement and first responders, and for sending members of our team on location. Our operations center also has rider support specialists, who can communicate directly with our passengers through our in-vehicle audio system.
Following a collision, we’re able to analyze all available data, including video and other sensor data, to evaluate factors that may have contributed to the incident, and we’re able to make any appropriate software changes and update every vehicle in our fleet accordingly. Any damage our vehicles sustain in a collision is repaired and the vehicles are tested for safety before they return to the road.
Last edited:
jimmy_d
Deep Learning Dork
I feel your pain. My mom's driving was scaring us enough that we had to restrict her to Uber. Which seemed like it was going to be a tough sell until she realized she could talk the ears off of the nice driver fellows and they were still polite.
Besides the snippet you quoted, I've skimmed that document and it's great! Definitely worth at least a skim from beginning to end.
Anyone who keeps thinking that level 4 or 5 automation (https://www.sae.org/misc/pdfs/automated_driving.pdf) will be available in the US to Tesla customers w/the right equipment within some of Elon's optimistic timelines might change their minds...
Last edited:
I guess Tesla is counting on their idea that "twice as good as a human" is good enough for the regulators. We shall see how that goes...
The Fisker E-motion will use 5 solid-state Quanergy S3 Lidar sensors.
Fisker EMotion On Track For CES Debut, Will Feature 5 LiDAR Sensors
Fisker EMotion On Track For CES Debut, Will Feature 5 LiDAR Sensors
The rest of the car looks fine, but I agree the front looks horrible. IMO, that was the issue with most of Fisker's designs (previously it was the mustache).
The next big battle in the autowars will not be fought over clay models or paint choices.
It will be pure tech. An AV only requires Prius level performance, and higher performance works against anti-collision systems.
Will people buy an AV that looks like an Edsel? You betcha. Lots of goofy looking high tech has been historically dominant. Heck, all desktop computers looked like shipping boxes painted with file cabinet paint. Only the Mac looked fashionable. And we ALL know what happened there.
When a working AV hits the showrooms, it will be 'shut up and take my money', not 'does it come in free-range tofu dash trim?'
It will be pure tech. An AV only requires Prius level performance, and higher performance works against anti-collision systems.
Will people buy an AV that looks like an Edsel? You betcha. Lots of goofy looking high tech has been historically dominant. Heck, all desktop computers looked like shipping boxes painted with file cabinet paint. Only the Mac looked fashionable. And we ALL know what happened there.
When a working AV hits the showrooms, it will be 'shut up and take my money', not 'does it come in free-range tofu dash trim?'
That's only temporary I suspect, they have an internal mapping project called Vahalla based on open source mapping and routing engines that is a drop in for the current Navigon stuff, it's deployed in the firmware now, they just haven't turned it on.
The automakers don't seem to believe that though. They still think styling matters a lot, which is why almost all manufacturers still insist on plenty of fake grill area on the front fascia and also on hiding sensors using various tricks (for example a blacked out area of the logo).
I think desktop computers most people don't care because they have the computer itself in some place they don't see often (behind a shelf or under the desk).
Simple really: don't allow Level 5 unless it's significantly better than humans. You can get very large gains with good Level 3.
Doesn't need infrastructural changes. So much congestion is caused by the failings of humans behind the wheel that Level 5 autonomy should significantly improve traffic flow with reduced collisions, improved reaction time, consistency and unselfishness. Better flow will mean improved efficiency. If it makes things worse it'd be because if it adds vehicle miles.
One I built for my son (2010):
It's still running, looks even better, and he always gets compliments on it. And it's pretty fast even today, OC'd 4.4g? Xeon ES on water w/GTX1080 w/RAID0 SSD's.
(non-geek translation) - CPU is running 1/3 faster than advertised using an Engineering Sample 6 core server chip by using water cooling, runs a very fast graphics card, and runs 2 solid state hard drives that run in parallel for near double the access speed)
Last edited:
Nissan is testing their next generation L4 capable propilot in Tokyo which they hope to release in 2020:
Nissan tests fully autonomous ProPilot tech on Tokyo roads
Nissan tests fully autonomous ProPilot tech on Tokyo roads
Similar threads
- Article
