Slashdot Mirror
IBM's Patent-Pending Traffic Lights Stop Car Engines
theodp writes "I'm sorry Dave, I'm afraid I can't let your engine idle. The USPTO has just published IBM's patent application for a 'System and Method for Controlling Vehicle Engine Running State at Busy Intersections for Increased Fuel Consumption Efficiency.' Here's how Big Blue explains the invention: 'The present disclosure is directed to a method for managing engines in response to a traffic signal. The method may comprise establishing communications with participating vehicles; responding to a stop status indicated by the traffic signal, further comprising: receiving a position data from each participating vehicles; determining a queue of participating vehicles stopped at the traffic signal; determining a remaining duration of the stop status; sending a stop-engine notification to the list of participating vehicles stopped at the traffic signal when the remaining duration is greater than a threshold of time; responding to a proceed status indicated by the traffic signal, further comprising: sending a start-engine notification to a first vehicle in the queue; calculating an optimal time for an engine of a second vehicle in the queue to start; and sending the start-engine notification to the second vehicle at the optimal time.' IBM notes that 'traffic signals may include, but are not limited to, traffic lights at intersections, railway crossing signals, or other devices for indicating correct moments to stop and to proceed.'"
Hmmm, a computer at a railway crossing that can remotely disable a car's engine. To use the parlance of our times "What could possibly go wrong?"
...if you don't ignore the fact that this is a blatant case of "patenting the goal". The patent is "here's a bunch of ideas that might work to control fuel consumption at signals, we claim them all."
VW (and other car manufacturers probably) already have cars that shut off at stop lights. The "3L" car they made (78.4 MPG, no batteries required) shuts down at stop lights. All this is doing is making it 'intelligent'.
This patent would be much less necessary if cities would install intelligent traffic lights that allowed traffic to flow and thus minimized idling engines.
"We can categorically state we have not released man-eating badgers into the area." - UK military spokesman, July 2007
This is a great idea. But it's being done the wrong way. Rather than sending signals to start and stop cars, the traffic lights should just send signals saying how long the red light is going to be, and while they're at it, specify how long the green lights will be too. Then the smart cars should interpret that how they will, by stopping, starting, or showing a light to the driver. This method will upgrade smoother and be more resistant to jokers with toys at the intersections.
suppose your car has been told to shut off at a red light. What happens if you need to suddenly perform crash avoidance? One of the standard things taught in driver school is leaving enough room between you and the car in front of you in case you need to avoid a rear collision. I'm not sure insurance companies are going to go for this.
...if you don't ignore the fact that this is a blatant case of "patenting the goal". The patent is "here's a bunch of ideas that might work to control fuel consumption at signals, we claim them all."
I agree that that this isn't really a fully-disclosed invention. Generally, IBM is more interested in patenting as much as possible just to create patent thickets and later shut out or tax real innovators with bullying tactics. The blog post I just linked to also mentions IBM's claim (made in early 2009) to have a number of patents "larger than those from Microsoft, HP, Oracle, Apple, EMC, Accenture, and Google combined." The blog post also mentions research that shows the average commercial value of an IBM patent is fairly low as compared to the portfolios of such competitors as Microsoft. The patent that gave rise to this slashdot article may be another example.
IBM has also been a long-standing aggressive force in pushing the envelope concerning the scope of patentable subject matter in the field of software. Courts can't be lobbied the way politicians are lobbied (which is something at which IBM is also extremely aggressive) but companies can try to bring up court case after court case pushing the envelope with new arguments in order to find loopholes to extend the range of what's patentable. The recent landmark decision in Germany, effectively lowering the bar for software patentability in the largest EU member state, was related to a Siemens patent, but other landmark cases in the US as well as in Europe (at the level of the European Patent Office as well as in individual European countries such as Germany) related to legal recourse sought by IBM in order to obtain patents on "inventions" of an ever lower standard.
At the lobbying front, the FFII (a European non-governmental organization fighting software patents and pushing for open standards) listed IBM as one of the four IT companies pushing hard for an overall patent and patent court reform in Europe aiming to strengthen the rights of patent holders and the legal basis for software patents.
This doesn't mean to say that IBM is the only company doing it, let alone the only one with an interest in this, but others entered the game relatively late and IBM has a history of decades of pursuing that agenda of an ever broader scope of patentable subject matter.
I wish all funding for that would be diverted to making a car with a survival instinct: Proximity sensors for collision avoidance, sensors to determine road conditions of maximum safe speeds accordingly, etc.
Once it becomes rare for someone to die in a car accident, THEN they can mess around with red light idling algorithms and self-driving cars. Just pick your priorities: Safety first.
You can't take the sky from me...
Upon receiving the stop-engine notification, the vehicles may automatically switch off the engine, or display an alert informing drivers to manually switch off the engines. A vehicle may optionally notify the service once its engine is switched off.
For all of you concerned about not having complete control.
The summary doesn't effectively explain when this would be useful. At most lights, it won't matter. The example the patent gives is a 2 minute light, for which it is inefficient to restart the engine state. It suggests "waiting for 10 minutes for a railway to clear" as a case where this would be useful.
The patent seems very vague. It talks about processing information about the movement about other cars, doesn't comment on what should be looked for, how that information is to be determined, or how wait-time should be estimated. It basically seems to be a patent for the idea of signaling the car when a long wait is anticipated.
What this patent fails to account for is that starting up the car results in increased fuel consumption for the short period while the engine attains running speed. Short period, though, but multiply it by the number of signals in an average city, and it might just come out that this actually increases sum consumption.
Also, I'd like to draw your attention to a post detailing just what can happen if we introduce networking into cars. And this is even made easier by the forced standards needed for this project to work...
Hyperbole: I use it liberally!
Anyone want to take bets on how long until the protocols gets hacked and spoofed?
This should be a good thread on comp.risks.
In a lot of busy roundabouts you still need traffic lights, otherwise you can have one flow of traffic blocking everyone else out.
In 20 years of driving, only the emergency vehicle one. I've never had any of those other situations. That's not to say they don't happen, but I am left wondering why you imagine they are common occurrences that happen to everyone.
Secondly, I wonder why whenever a new idea/patent/invention is brought up here, some people object to it on the basis of assuming that the worst possible implementation is the one that would be done; that the inventor hasn't already considered the problems that come into your mind within seconds of hearing the idea, and dealt with them. And that an implementation would be continued with if such problems remained.
a sizable number of people already do switch off their engines if they are in a quene of traffic and are going to be waiting a long time. And this doesn't seem t have caused problems. So there doesn't seem to be a problem with switching engines off in queues per se.
A rational implementation would of course not switch the engine off unless the car was already stationary with the parking break applied. And it would not stop the driver from restarting the engine with the ignition key if needed. These obvious details seem to deal with all your fears.
The ability to turn cars on and off at their whim.
How long until one hacks into the system and just turns them all off?
When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
All this is necessary to achieve this is wider application of so called "mild hybrid" technology. Think over-sized battery, over-sized starter motor, over-sized alternator, drive by wire throttle, and a bit of ECU smarts. Any time the car is stopped (or below speed X, where X is small 3mph? ) the engine is "off" (no fuel, no spark, engine still turning if the car is in motion), and the battery and starter motor move the car. Once the speed threshold is exceeded (or battery is sufficiently low) fuel and spark are resumed. At any greater speed the vehicle is powered entirely by the engine. Having electrically powered accessories (power steering assist, air conditioning, brake booster, etc.) would be preferable, so that all of those systems still function when the engine is not spinning. It addition, making these electric tends to increase efficiency and reduce weight. If these systems remain belt or vacuum driven there can be many situation where the ECU may be forced to leave the car idling, or waste battery power spinning the engine to keep the accessories running.
pending adj (postpositive) 1. not yet decided, confirmed, or finished
While it is true that the patent system is completely broken, you really should cut the USPTO some slack here. The patent was only applied for on Thursday. It hasn't even been looked at yet!
Published Thursday. Filed November 14, 2008. But still probably not examined.
How about first they invent a 60-second kill switch on the dome light so that when your children leave the dome light on all night, and then jump the car in the morning, you don't have to rev the engine at every stoplight to have a hope of restarting the car after leaving the destination?
First of all, starters have a limited lifetime. If you force cars to engage them at nearly every stoplight, they will wear out 10 times faster or more. If a starter wears out at a traffic light, the car can't start and the flow of traffic will stop. This will do wonders for fuel savings, but not much good for transportation, as a whole.
Second, if my engine shut off at a stoplight in the Texas summer, my air conditioning would not work and I would effectively be baking in an oven.
How about instead of that a marquee that tells motorists the wait time and suggests shutting off their engines to save gas?
I'm betting it will be cheaper (especially since you can't patent a sign saying how long the wait is, theme parks have decades of prior art), it completely avoids an entire class of serious problems (like what if a carjacker/terrorist/bored 13 year old figures out how to send the kill signal) and doesn't require refitting every car in the U.S.
If voluntary participation isn't widespread enough once the signs are in place, put on a few PSAs encouraging kids to pester their parents about it.
Of course the whole thing will become a moot point anyway as hybrid and electric cars take over. At that point, at least my alternative suggestion can then be re-purposed to provide news and entertainment.
Detroit was killed by the UAW, not by anything that has anything remotely to do with the race of the citizens.
"Workers to power!"
That doesn't "work" very well in practice, and couldn't even compete with the Capitalism you hate in terms of worker benefits. Eating the rich is great fun until you run out of rich to eat, then everyone else goes on the exploitation menu, for the good of the workers, of course.
Capitalism leavened with a humane but not overpowering dose of Socialism arguably produces the best results. Business should be restrained by government, government restrained by business, and both restrained by the votes of an engaged and informed citizenry. Makes a lousy slogan though.
"This post is an artistic work of fiction and falsehood. Only a fool would take anything posted here as fact."
Stopping cars at the light isn't all that hard, but making it a good idea to stop them at the light is pretty far away. The trivial problem of pre-loading the oil pump so that it does not do serious life limiting of your engine is a small detail that nobody seems to care about (because they don't know that starting and stopping conventional vehicles at every intersection is murder to the lubrication system, and therefore the entire engine.) Some cars do this (prius, some police cruisers) but most do not.
Much more, you have to re-engineer cars so that A/C is not dependent on the motor running.
Many people who live in places without shit weather have no idea about this (particularly California where "green" ideas that are wildly impractical seem to come from).
Are there places in the country where you can freeze to death if your car stops? yes. Are there places in the country where infants can die of heat stroke if your car stops? Yes. those places just don't happen to be in California/NY.
Heat and A/C are not about comfort, they are about survival, at least in many places in the country.
Let's see...
"suddenly having to run the light to avoid getting rear-ended by someone coming up behind you completely oblivious to the light?"
Now I've certainly seen rear end collisions at traffic controlled junctions. But the evasion manoeuvre described requires all of the following:
1) That as you are approaching a junction you are watching the rear view mirror.
2) That you are the first vehicle that would otherwise be stopped by the traffic lights. (You can't outpace a collision from behind if there is a stationary car in front of you.)
3) i) That you are in motion as a speed that you COULD reasonably stop, and yet know that the car behind you is at a speed that it could not stop.
Or ii) That your car has dragster like acceleration to get you moving from stationary and across the junction fast enough to avoid both collisions from behind, and potential collisions from the side.
* 3i does not seem to be relevant, since the idea is a mechanism to save fuel and emissions whilst a car is stationary, not an idea to slow a car to a halt by stopping the engine whilst the car is still moving.
All in all there is very little chance this has happened to either of you. Rather what we have here is the perfect demonstration of the tendency of the male of the species to talk absolute bollocks when it comes to the topic of driving experiences.
The experience that I do believe you have had is that you've run a red light yourself, thought you cut it a bit fine, and then noticed that another car behind you has also jumped the red light. Easy enough for the self justifying driver to kid himself about the reality of that situation.
I have encountered traffic lights which never turn green for my direction, even after waiting through many cycles for about 5 or 10 minutes. There was one light in town which seemed to be wait for its buried magnetic sensing loops to detect at least two cars to arrive before it would turn green. I would be coming back from a grocery store at about midnight and have to sit there for about 5 minutes while waiting for another car to arrive, then it would finally change. Unfortunately, there was also a no right turn sign at the intersection, which made it impossible for me to legally escape from the situation.
There were three lights in town, more or less, like that. I have wondered if perhaps something about my pickup truck or the way that I approach the intersection, fails to trigger the buried magnetic sensing loops. I once talked to another local resident who had the same problem, in his pickup truck, with one of the same traffic lights.
I would hate to encounter a traffic light which could turn off my engine in that situation, preventing me from backing up or turning right to escape the problem. Would there be an emergency override switch for the system? Would they require such a system being retrofitted to my older 1992 pickup truck? I also wonder what would happen if the power to the traffic lights goes out during a thunder storm. Would our engines default to start in such a situation, so that we could treat the failed traffic lights as 4-way stop signs?
When will manufacturers, especially software manufacturers, ever understand the concept that it is *MY* computer or device, *NOT THEIRS* ???
As noted above in all the "What could go wrong?" posts, this kind of central control is fraught with problems and unintended consequences..
If they simply take an approach to design and engineering that respects fact that it is not their device, all kinds of problems go away.
A proper approach would be for the lights to broadcast their status and schedule for the next few minutes (i.e., how long until the next change, how long will be the next red, etc.), and allow the vehicle and driver to decide what to do about it.
Sure, If we're at the beginning of a long red, then it is probably best to shut down. But, if we're making a right turn and/or trying to get someone to the hospital at 3AM, have paused to check that there is no crossing traffic, then we should drive on. If the hybrid motor is trying to recharge low batteries, the motor should keep running. Etc. We could even have a dashboard or heads-up display showing the status so the driver can make better decisions. Different car designers can code the best algorithm for *their* particular car design, e.g., a hybrid might use a completely different response pattern than a truck or a sportscar.
What is so hard about that? [Warning - oversimplification following] Decentralized systems are generally more flexible, and have shallower bugs than centralized systems. So, why do they persist in designing that way?
... why not sync the %$#@&)$$ lights better? That would save quite a bit of gas, not having to start and stop at every intersection.
Have gnu, will travel.
IMHO the biggest problem with traffic lights isn't synchronization; it's the behavior of the drivers. At least in my area the lights are attached to inductive sensors placed in the tarmac, and the way they're supposed to work is that they sense the presence of the car approaching the light and, after a suitable period of wait time hysteresis (which starts from the last time the light switched, and so may already be expired), switch the signals and allow the car to pass.
Unfortunately, drivers (again, at least in my area) aren't very clueful about the presence of the sensors, and will stop way, way back of the stop bar, before they get to the sensor, or pass over it and stop halfway into the intersection. The sensors are huge, roughly 2m by 8m, so it's not like you have to be precise to hit them, and they are visible as grooved loops in the tarmac just behind the stop bar, but I can't count the number of times I've been stuck behind a long line of cars at a light, with the first car stopped before it got to the sensor. As far as the light can tell, there's nobody stopped at the light, so we wait and wait.
I've driven with people who have stopped before the sensor and then complained about how poorly the lights are "synchronized." Apparently, traffic light sensors are not common knowledge.
Why would you need to shut off your car when making a delivery? It only takes a minute to hand the pizza, take the money, and go back to your car. Every pizza guy I've had leaves the car running.
Though I dont know about the OP, I have delivered pizzas in areas, that though not "slums" or "crime ridden" it still would be a bad idea to leave a car running while running into a townhouse/small apartment complex to deliver a pizza; if only for the simple fact that some delinquent may decide to take it for a joy ride.
Places like the Baltimore suburbs (and increasingly Long Island) have a lot of 2 story, townhouse-looking apartments that a pizza delivery to requires going up a set of stairs or into an enclosed vestibule where your car is out of your sight.
StarTrekPhase2 - The Five Year Mission Continues!
Ride a motorcycle for a few years before you question the scenario. Any long term motorcycle rider will tell you that he's spot on. You either learn how to "check your six" at stoplights and be prepared for evasive maneuvers at maximum speed or one day you _will_ have an automobile up your ass, literally.
At stoplights I commonly am watching the rear view mirrors and I'm sitting with the clutch pulled in and the bike in gear so if I see something all I have to do is loosen the grip with my left hand and roll the throttle with my right.
Would you care to guess how many times that's saved me from being rear-ended at a stop sign / light?