"All cyclists will need to fit detection beacons, says cycle industry boss"
#101
Senior Member
Join Date: Oct 2015
Posts: 15,481
Bikes: 2015 Workswell 066, 2017 Workswell 093, 2014 Dawes Sheila, 1983 Cannondale 500, 1984 Raleigh Olympian, 2007 Cannondale Rize 4, 2017 Fuji Sportif 1 LE
Mentioned: 144 Post(s)
Tagged: 0 Thread(s)
Quoted: 7649 Post(s)
Liked 3,465 Times
in
1,831 Posts
This. And since he doesn't understand what a troll is he likely doesn't even know he is one. A debate is not insults and name calling, but requires some level of logic and direction leading towards a conclusion. A troll has no purpose except to create conflict and discord.
"Argument is an intellectual process. Contradiction is just the automatic gainsaying of anything the other person says."
https://www.bing.com/videos/search?q...BF92&FORM=VIRE
#102
Senior Member
Join Date: Jun 2013
Location: 700 Ft. above sea level.
Posts: 3,236
Bikes: More than there were awhile ago.
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 629 Post(s)
Liked 1,258 Times
in
602 Posts
"An argument is a connected series of statements intended to establish a proposition.
"Argument is an intellectual process. Contradiction is just the automatic gainsaying of anything the other person says."
https://www.bing.com/videos/search?q...BF92&FORM=VIRE
"Argument is an intellectual process. Contradiction is just the automatic gainsaying of anything the other person says."
https://www.bing.com/videos/search?q...BF92&FORM=VIRE
__________________
".....distasteful and easily triggered."
".....distasteful and easily triggered."
#103
Banned
Join Date: Apr 2010
Location: Lincoln Ne
Posts: 9,924
Bikes: RANS Stratus TerraTrike Tour II
Mentioned: 46 Post(s)
Tagged: 0 Thread(s)
Quoted: 3352 Post(s)
Liked 1,056 Times
in
635 Posts
As always I am amazed by what some people can think up to control other people. IMO it must be some kind of birth defect.
#104
☢
#105
What happened?
Join Date: Jun 2007
Location: Around here somewhere
Posts: 7,927
Bikes: 3 Rollfasts, 3 Schwinns, a Shelby and a Higgins Flightliner in a pear tree!
Mentioned: 57 Post(s)
Tagged: 1 Thread(s)
Quoted: 1835 Post(s)
Liked 292 Times
in
255 Posts
SR-22 will file your insurance proof claim over a mobile device? (Look up SR-22)
__________________
I don't know nothing, and I memorized it in school and got this here paper I'm proud of to show it.
#106
Senior Member
Join Date: Apr 2005
Location: Ann Arbor, MI
Posts: 16,866
Bikes: 1980 Masi, 1984 Mondonico, 1984 Trek 610, 1980 Woodrup Giro, 2005 Mondonico Futura Leggera ELOS, 1967 PX10E, 1971 Peugeot UO-8
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 1854 Post(s)
Liked 661 Times
in
504 Posts
What else do we put these devices on? Kid's bikes? Mobility scooters? Push chairs? Skips? Traffic cones? Cat/dog collars? Sheep...?
Is it not right that any moving vehicle should be able to prevent hitting a universal object in it's path..? That's what cars have drivers for - if you are going to remove the driver, the car will have to be able to execute that same function regardless of what that universal object is.
Is it not right that any moving vehicle should be able to prevent hitting a universal object in it's path..? That's what cars have drivers for - if you are going to remove the driver, the car will have to be able to execute that same function regardless of what that universal object is.
Actually, independent of the AV development there has been thinking about beaconing bridge abutments, constrution barriers, and other things that a car or a driver might bump into at speed.
#107
Senior Member
Join Date: Apr 2005
Location: Ann Arbor, MI
Posts: 16,866
Bikes: 1980 Masi, 1984 Mondonico, 1984 Trek 610, 1980 Woodrup Giro, 2005 Mondonico Futura Leggera ELOS, 1967 PX10E, 1971 Peugeot UO-8
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 1854 Post(s)
Liked 661 Times
in
504 Posts
The reactions here and on the original blog story are largely talking about AVs. The whole issue has history, going back 15 to 20 years, before AVs were seriously considered a possibility.
Today many cars have Forward Vehicle Collision Warning functions that have trouble detecting cycles and pedestrians. At the time they were designed, the attitude was to proceed with overall collision prevention (aka traffic fatality prevention) by taking the baby steps that technology could enable. The OEMs and sensor companies knew that radar could detect cars and trucks and had trouble with smaller pieces of metal or carbon, humans, and large and small animals (from moose on down). Hence the qualification "Vehicle Collision." Based on the philosophy "it is more morally good to prevent the fatalities that you can prevent, than to not prevent them because you can't prevent them all," standards for such feasible car features were written at ISO and SAE, and carmakers and their suppliers implemented such systems using those standards as models. 15 years ago, these systems had nothing to do with AVs. Beaconing was recognized as a possibility, but good options did not readily exist for beaconing people walking or pedaling.
Additionally, the annual US automotive fatality stats showed several tens of thousands of fatalities due to vehicle-vehicle collisions, and less than 1000 for cyclists. It's pretty straightforward to show the car-car data was pretty comprehesive, but the car-cyclist data is less certain, at least for non-fatal injury collisions (not to say that any such data set is faultless!).
Beacons (microwave DSRC using ODFM or cellular V2x) came out as a concept when it was realized the vehicle external sensors should feed external objects to an in-car database, and that the database could contain walking or cycling people on or near the roadway. If consumers would accept such a solution and system could be designed to be managable across the many components handling data, cars could tell their drivers to take action to avoid hitting pedestrians and cyclists.
Hence the beacons. It is a really difficult sensor problem, to reliably detect and track (without the tracking you cannot calculate a pre-collision path) small metallic objects and non-metalic objects in the presence of many large metallic objects. In other words, the presence of many cars, lampposts, sign poles, and other urban items can partly mask the signals reflected from people and bicycles - in RADAR technology it's called clutter. If a beacon can provide my (as a cyclist) GPS position well enough to infer speed and direction of motion to the database of external object tracks, the car can help its human driver avoid hitting me with more reliability than the Mark 1 eyeballs.
Today many cars have Forward Vehicle Collision Warning functions that have trouble detecting cycles and pedestrians. At the time they were designed, the attitude was to proceed with overall collision prevention (aka traffic fatality prevention) by taking the baby steps that technology could enable. The OEMs and sensor companies knew that radar could detect cars and trucks and had trouble with smaller pieces of metal or carbon, humans, and large and small animals (from moose on down). Hence the qualification "Vehicle Collision." Based on the philosophy "it is more morally good to prevent the fatalities that you can prevent, than to not prevent them because you can't prevent them all," standards for such feasible car features were written at ISO and SAE, and carmakers and their suppliers implemented such systems using those standards as models. 15 years ago, these systems had nothing to do with AVs. Beaconing was recognized as a possibility, but good options did not readily exist for beaconing people walking or pedaling.
Additionally, the annual US automotive fatality stats showed several tens of thousands of fatalities due to vehicle-vehicle collisions, and less than 1000 for cyclists. It's pretty straightforward to show the car-car data was pretty comprehesive, but the car-cyclist data is less certain, at least for non-fatal injury collisions (not to say that any such data set is faultless!).
Beacons (microwave DSRC using ODFM or cellular V2x) came out as a concept when it was realized the vehicle external sensors should feed external objects to an in-car database, and that the database could contain walking or cycling people on or near the roadway. If consumers would accept such a solution and system could be designed to be managable across the many components handling data, cars could tell their drivers to take action to avoid hitting pedestrians and cyclists.
Hence the beacons. It is a really difficult sensor problem, to reliably detect and track (without the tracking you cannot calculate a pre-collision path) small metallic objects and non-metalic objects in the presence of many large metallic objects. In other words, the presence of many cars, lampposts, sign poles, and other urban items can partly mask the signals reflected from people and bicycles - in RADAR technology it's called clutter. If a beacon can provide my (as a cyclist) GPS position well enough to infer speed and direction of motion to the database of external object tracks, the car can help its human driver avoid hitting me with more reliability than the Mark 1 eyeballs.
Last edited by Road Fan; 04-02-18 at 08:33 AM.
#108
Cycle Year Round
You did watch how the Uber safety driver handled the situation in the video, correct?
__________________
Land of the Free, Because of the Brave.
Land of the Free, Because of the Brave.
#109
Senior Member
Join Date: May 2008
Location: Zang's Spur, CO
Posts: 9,080
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 3370 Post(s)
Liked 5,491 Times
in
2,844 Posts
I have never had one fail to detect me on a bicycle.
#110
Senior Member
Join Date: May 2008
Location: Zang's Spur, CO
Posts: 9,080
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 3370 Post(s)
Liked 5,491 Times
in
2,844 Posts
Another Tesla owner reproduced the situation at the location of the recent fatality.
Watching the LCD display indicates the car loses reference of where the righthand lane markings are at 0:28.
The "gore point" painted on the pavement is missing the white line that should mark the right side of it.
It appears the car was using the stripe along the left side of the gore point as it's left lane edge reference.
Now why the car's radar did not detect the barrier is an open question.
Watching the LCD display indicates the car loses reference of where the righthand lane markings are at 0:28.
The "gore point" painted on the pavement is missing the white line that should mark the right side of it.
It appears the car was using the stripe along the left side of the gore point as it's left lane edge reference.
Now why the car's radar did not detect the barrier is an open question.
#111
Cycle Year Round
#112
Senior Member
Join Date: Jun 2008
Location: Louisville, KY
Posts: 13,443
Mentioned: 33 Post(s)
Tagged: 0 Thread(s)
Quoted: 4224 Post(s)
Liked 2,944 Times
in
1,803 Posts
#113
Senior Member
Join Date: Jun 2008
Location: Louisville, KY
Posts: 13,443
Mentioned: 33 Post(s)
Tagged: 0 Thread(s)
Quoted: 4224 Post(s)
Liked 2,944 Times
in
1,803 Posts
#115
Senior Member
Join Date: Apr 2005
Location: Ann Arbor, MI
Posts: 16,866
Bikes: 1980 Masi, 1984 Mondonico, 1984 Trek 610, 1980 Woodrup Giro, 2005 Mondonico Futura Leggera ELOS, 1967 PX10E, 1971 Peugeot UO-8
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 1854 Post(s)
Liked 661 Times
in
504 Posts
#116
Senior Member
Join Date: Apr 2005
Location: Ann Arbor, MI
Posts: 16,866
Bikes: 1980 Masi, 1984 Mondonico, 1984 Trek 610, 1980 Woodrup Giro, 2005 Mondonico Futura Leggera ELOS, 1967 PX10E, 1971 Peugeot UO-8
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 1854 Post(s)
Liked 661 Times
in
504 Posts
And when the human driver stops using the Mark 1 eyeballs because the car will warn them and you happen to be cycling in a tunnel or heavy forest with no or poor GPS, what happens to the cyclist then?
You did watch how the Uber safety driver handled the situation in the video, correct?
You did watch how the Uber safety driver handled the situation in the video, correct?
The company I worked for developed some of the Warning systems I described and we expanded some of them into an automatic emergency braking system. When we put together a fleet of cars equipped with them, we also equipped them with a camera that watched the driver's face and which had an algorithm that would detect when the driver was inattentive. There were other measures to evaluate the test drivers, as well. It could then take action to alert the driver, such as a sound, pulsing the brake pedal, vibrating the driver seat. An OEM (Original Equipment Manufacturer or traditional car company) who contracted us for our driver sensing system (which I actually developed) was free to use however raucous a sensory warning as they saw fit. Clearly Uber did not have such a function active in the vehicle that was being tested.
You recall how Uber had switched off the AV system provided to them by Volvo as part of the car's equipment? We used the same sensing technology as Volvo does in their cars. This further implicates Uber. They should have had the sense to cherry pick which Volvo features they were not going to use, and learned something.
I think the multi-functional driver safety assistance systems on the market from the traditional OEMs (i.e. Ford, GM, Toyota, Subaru, Mercedes and many others) are rather well thought out, compared to what we're seeing on the so-called AV side from the technological world - this includes Tesla IMO. I am impressed at how little such system failure/collision news is apparent from Waymo.
Last edited by Road Fan; 04-02-18 at 08:44 PM.
#117
Cycle Year Round
CB: Yes, I did. I think the car should have slapped or shocked her, recorded and time stamped her responses, and then fired her when returning for work. In my opinion that behavior was completely irresponsible when monitoring the behavior and effectiveness of the so-called "autonomous" system, which appears to be not proven nearly as rigorously as I've come to expect in the auto industry. In commercial aircraft the vehicle monitoring task (what the cockpit crew does continuously to make sure the auto pilot isn't flying them into a mountainside) is regarded as very taxing, and that's for trained, talented, skilled, certified, and experienced air pilots who know the importance of what they are needed to do. Due to increased clutter and lax protocols I'd say the automotive task is more challenging.
https://www.seattletimes.com/life/tr...oting-airport/
__________________
Land of the Free, Because of the Brave.
Land of the Free, Because of the Brave.
#119
Cycle Year Round
#120
Senior Member
Join Date: Oct 2015
Posts: 15,481
Bikes: 2015 Workswell 066, 2017 Workswell 093, 2014 Dawes Sheila, 1983 Cannondale 500, 1984 Raleigh Olympian, 2007 Cannondale Rize 4, 2017 Fuji Sportif 1 LE
Mentioned: 144 Post(s)
Tagged: 0 Thread(s)
Quoted: 7649 Post(s)
Liked 3,465 Times
in
1,831 Posts
#121
☢
Most new cars already have (standard/optional) pedestrian detect technology. No beacons necessary.
#122
Tortoise Wins by a Hare!
Join Date: Feb 2006
Location: Looney Tunes, IL
Posts: 7,398
Bikes: Wabi Special FG, Raleigh Roper, Nashbar AL-1, Miyata One Hundred, '70 Schwinn Lemonator and More!!
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 1549 Post(s)
Liked 941 Times
in
504 Posts
#124
Senior Member
Join Date: May 2008
Location: Zang's Spur, CO
Posts: 9,080
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 3370 Post(s)
Liked 5,491 Times
in
2,844 Posts
It does not calculate range, azimuth, or elevation, nor does it try to track multiple targets.
It only detects the speed of the strongest return signal.
So if a vehicle behind you is giving a stronger return, that is the one it will display.