Vehicle Tracking

Interesting idea. I don't know how much the hose sensors cost. Maybe a good place would be winding mountain roads that are not heavily traveled but have a reputation for off-road accidents.

You still have to have some way of knowing how many cars are between the count up and count down sensors, which depends on the traffic density.

I don't think the hose type sensors will be durable enough especially on interstates.

I am thinking a photo reflector or emitter-receiver pair.

The pair would be separated by at least several feet to prevent false triggering by leaves,animals,etc.

Use a multiple wavelength source to minimize signal loss in inclemate weather.

It may even be possible to use a sonar-type beam,like in level controllers,or a combination of both methods.

I think hall effect sensors will be too expensive to install,as well as having a more complicated circuitry to require calibration and maintenance.

Configure the chosen detectors to be hot swap-able to simplify replacement.

Mount these sensors on the shoulder or in the median.

Offset one side of the road from the other,and you will have an absolute-type encoder to indicate vehicle direction.

This data would be wirelessly transmitted to a CPU for tracking purposes.

Cell towers could be used as a relay to a centralized facility.

The main information you will be looking for is a large discrepancy between the sensors.If 5,10 or 20 cars do not show up at the next sensor,you have a problem.

All traffic has probably stopped.!!!!

On two lane roads,there should always be a gap between vehicles,unless someone is passing.

On interstates,I can see both lanes being fully occupied,but there will still be some gap between vehicles unless there is a problem.

A retro-reflective photo sensor on each side of the interstate (median and shoulder)

with a certain response distance will indicate doubling of vehicles.

The absolute number is not as important as the group numbers.

I think you are basically looking for traffic flow sensors. Different devices in use now are described in this document.

https://www.fhwa.dot.gov/publications/research/operations/its/06108/01.cfm

That concept was similar when Paul Mollor was development the flying car in the 70’s??… for traffic control.

But at the time, the technology wasn’t there yet

Why not cameras with id tracking and AI controller with drone launch capabilities to search and locate possible missing vehicles...?

Feels a little too big brothery....

Maybe you could just expand on the on-star system that notifies 911 when an accident is detected...?

All along I-81 in Virginia, the VDOT folks installed cameras that would appear to have nearly continuous coverage of the highway. Big Brother IS watching you while the W.H. censors you. Did I say that?

https://www.511virginia.org/

With machine vision algorithms, something unusual could flag a human to have a look and determine if an emergency response is warranted. With only two lanes in each direction for most of the interstate, even small accidents cause massive back-ups on I-81. And given the very high volume of truck traffic and the many hills along this route, give ample opportunity for people to do something stupid.

Cameras are probably a lot less hardware intensive than other sensor systems.

I agree, I think the 511 system in operation in all states that I travel in handles most if not all these issues.

This might work on limited-access roads where vehicles enter and leave the road only at fixed locations. Parking lots, homes, side streets, and curbside parking along the road will make this impractical if not impossible to implement. Using directionless sensors that cannot differentiate if a vehicle is entering or leaving a roadway will compound problems, too.

The photoemitter-sensor with or without reflector can only work for a single lane or multiple lanes that have some form of separation between lanes, where the second device can be nounted. With two lanes moving in the same direction, a car tailgating another, with a third bridging the gap between them in the other lane could appear to be a single long vehicle. A car adjacent to another one in the other lane would only appear to be a single vehicle.

It's conceivable that having the photo beam at a fairly high angle above horizontal could solve this problem, with one device close to the ground and the other high enough that High trucks in the second lane wouldn't block the path.

There is no perfect system of any type.

Cameras require constant monitoring,usually with hundreds of monitors viewed by over worked employees.

Tracking objects is a mature technology.

It is used in airport terminals,warehouses,and postal sorting machines.

I am merely suggesting modifying an existing system to monitor traffic movement.

I am open minded to all suggestions for a perfect system.

Please suggest one.

A sonar transceiver,like used in level control systems could detect distance to target.Tis would detect vehicle at different distances-to- target,eliminating the gap-detection problem in most cases. There might even be a low-level sensor looking under the frames to detect the wheels on the cars.

With proper algorithms it could resolve the distance between wheels,like a semi,or a car.Unless the tires were perfectly aligned from one vehicle to the next,it would work.

Insofar as direction of travel,see my previous post (#2) regarding that issue.

....the only traffic management system likely to get approved and installed is the one that identifies the vehicle and links the licence to automatically debit a bank account, linked to rules for usage, instantly adjusted to respond to road conditions.

...which in the course of time can be 'sold' on 'safety' but arbitrarily adjusted and tweaked by computer algorithms to maximise income....

In UK there are cameras in place behind the information pantries.. Where they are hard to damage etc. Some are in a box and include recording format. Others are just curb flow sensors.

All are connected to computer systems that send flags to humans if needed and have APR systems built in.

It has been asked several times that such systems be set to poll the plates for tracking of stolen vehicles etc.. But at this time they don't do that. Instead, a weekly download is sent from the highways agencies to the local police from the traffic camera and daily alerts are sent within minutes for accidents or such..plus police can request searches of the old files to track vehicles after the fact.

So, big brother is watching in their but privacy concerns means a lot of things that could be used are not used.

In fact.. An area where my daughter lives has twelve different CCTV systems in place and the police need warrants to see any of them.. All so the privacy of the criminals is kept safe..

But that is not as bad as some of the Caricom countries Where CCTV is not allowable for prosecution of a case and if you post videos of a criminal breaking the laws you could actually be sued for his privacy being broken..

https://www.nar.realtor/sites/default/files/documents/2018-NAR%20Surveillance-Survey-Update.pdf

With all the electronics they are putting on cars, I suspect it's only a matter of time before all cars are required to have ELTs, Electronic Locator Transmitters, similar to those required on aircraft.

NOAA - Search and Rescue Satellite Aided Tracking - Emergency Beacons

I’m pretty sure Google and others are doing this already with whatever active GPS / location chips in cars right now, cars, phones, computers. How else do they get red & yellow indication so quickly on just about any stretch of road you come across?

Just anonymously read the locations of everyone on the road, and over time, you know speed, and then can infer typical traffic problems from those vectors.

This is already feasible and in place.

In Australia, there are camera systems that monitor heavy vehicle traffic for speeding "point to point" cameras. These validate licence plates and then use those to monitor each vehicle. Fines issued for speeding. They have been trialing facial recognition to validate the actual driver of each vehicle for log book systems.

This same system is being "observed" to see whether it will also work for cars. Of course it will, but i suspect they are monitoring the cost benefit impact of implementing that.

Those cameras, plus the traffic light cameras also have pattern recognition for people using mobile phones and issuing fines with supporting images to offenders.

Orwell was only wrong by a few years!!

The facility exists already. Mapping applications linked to a travel direction function look at the direction and rate of change of mobile telephone handset signals to determine whether or not a road if flowing freely, reporting it as colo(u)red "sausages" along the route where there is slow-moving traffic.

Intelligent satellite navigation systems use these data to indicate to the user whether or not there is a better route to the destination given current congestion conditions.

I think with most drivers having cell phones, the recognition that there are issues on the roadway is much faster by drivers reporting or the affected drivers self-reporting rather than relying on a downstream sensor. Interchanges and intersections make the analysis you are referring to much more complicated.

I do think milepost should be signed much more frequently and drivers trained to know what those signs mean. But, then GPS capability on the cell phone could resolve an incident location more accurately and with reliability.

Tolling booths could have been used in a manner similar to the sensors being described. But, the data was too complex and had a wide range of interpretation.

Isolated and roads not often driven might be a place for the described sensors. But, then, if traffic is that low, is it worth monitoring?

The <...warning signs to approaching vehicles, and even to poor visibility or hazardous road conditions ahead; Fog, heavy snowfall, heavy rain, black ice, etc....> is the Mk.1 eyeball. That's why drivers are trained and licensed.

All that will change with the onset of autonomous vehicles.

There's an app for that...

https://www.theverge.com/2019/10/1/20892875/google-pixel-personal-safety-app-car-crash-detection