Concept for new HW/SW to combat bad driving

**RetroPlayer** · 04-10-2006, 03:27 AM

I have had this idea for a while, but another thread recently (in SW development) inspired me to bring it up again.

Intro:
Everyday, we see people do things wrong while driving, and we look the other way. I mean, what can we really do? Many times, we are guilty of the same things ourselves. What would we gain by confronting someone? If the situation escalates, you will likely be seen as the instigator and will get all the flack about it. People seem to have trouble regulating themselves, resulting in more and more laws (police state) that are needed to regulate your behavior for you. This really isn't the solution. More laws just make more criminals, and besides...people will continue to break the law when they have a reasonable assumption that they will not get caught. Sometimes, things aren't actually illegal, but are just plain wrong and dangerous. In fact, I would argue that by creating more laws, we are just further destroying people's incentive to self-regulate. Basically, "It's only illegal if you get caught." Wonderful concept isn't it? How many times have you heard that? Sounds like an invitation to a police state to me!

The Problem:
Actually, there are so many popular bad driving habits that are wreckless, dangerous, and all too common that I couldn't possibly list them here. You know them when you see them.

[example removed since it wasn't universe-friendly and apparently distracted from the purpose of this topic]

IMO, the main reason for the rapidly decaying self regulation is that too many of these bad behaviors are culturally accepted. Not to mention that we would much rather someone else deal with the unpleasantness. This not acceptable! If the problem were not already so prevelant, it would be enough just to confront and scold someone. They would know you would be backed up and everyone else would see what they did as wrong. I could go on and on about the social aspects this, but I believe deep down, everyone understands this already.

The Solution:

As I already stated, my preferred solution is definitely not creating more laws. Creating a bunch of snitches is also not honourable. In fact, I submit that in many cases laws are not needed at all and police are not needed. One of the most effective deterrents against bad behavior is simple embarassment. This is no. 1 reason crime is so little in tribes and small communities. People regulate themselves because of acceptance by thier peers. Even people that claim they "could care less what other people think" only mean they could care less what people think outside thier circle of influence. It's very simple really. In many cultures, it is believed that exile is a punishment far worse than death. It's no coincidence that these cultures usually suffer very little crime!

So, in conclusion, the solution is very simple. If it were possible to somehow bring shame and scorn to wrong-doers

from within thier circle of influence it would be an effective incentive to self-regulate themselves.

By recording our daily driving experiences and publishing edited video on public TV (which is free and everybody can use) or local internet forums, we may actually see results. It would also remain somewhat light-hearted in that it would be entertaining (people love to watch other people do stupid things and think they are so much better!)

The biggest problem is legality, of course. I am aware of several concerns, and have yet to balance them into the system. I invite discussion on this. However, the overall possibility for good is too great to not explore this option. "Pie in the sky" ideas would be insurance companies offering a discount for having this system in your car, rewarding you for activism in safe driving. The obvious one result I would NOT want to see in this is the police using it to scan plates for people with warrants/suspended licenses, etc.. I am VERY much against a police state and 'spying' on your neighbor. It would be great if it could remain a simple citizen-activist tool. Any other purpose only invites abuse of power, which is never in the common good. So, if it isn't clear already, I AM NOT advocating that at all (sorry, had to be said.)

The Concept:
A software/hardware system cosisting of:

1. Video cameras located around the vehicle. Positioned to capture the majority of the field of view around the vehicle. Low light, high resolution, and concealed are important specifications of the cameras.

2. A video interface system.
This is the meat and potatoes of the system-
This system will take in telemetric data from your vehicle and overlay it on the video. Information such as speed, indicators, steering position, seat-belt, and other data can be configurable for display.
Additionally, key vehicle data is recorded seperately and timestamped in machine readable form regardless of what is overlaid on the video.
Additional features might include x minute looping of video capturing, while monitoring things like collision sensors and sudden loss of speed (braked or no brake.) A triggered event would store all related information seperately for later investigation. There are usually more than enough of these sensors built directly into a modern vehicle that installation would be easier and modifications minimal.

4. Software interface - depending on whether this is a standalone device or a background application to an existing carpc, this part may be different. The software should run entirely of in the background and require no (or minimal) input from the vehicle occupants while in operation. It should be fully configurable except where critical data is concerned. The software would be somewhat complex in that it loops the data of all the senors and all the cameras. A trigger event (either manual or from the vehicle's sensors) would store x number of minutes before and after the trigger. This is the hard part, actually. How easy would it be to remember to trigger the system manually? Of course most drives are fairly short and a trigger could be set using the ignition of the car, so it starts recording when turned on, and permanently saves the data when turned off (or prompting whether to keep the data, defaulting as "yes" with a timeout.)

A simple and convenient method of offloading the data should be available.

Conclusion:

I truly believe this has a great potential of working. It should be cheap enough for a normal citizen activist to implement. Simple enough for most people to use. Have safeguards against abuse of the technology. I know there are already commercial systems which provide many of the features I presented in CONCEPT, but they usually do not meet the above criteria.

It could only be truly effective if enough people are doing it, since it would be easy to figure out who was participating, especially when most people travel the same routes every single day. The information needs to be random enough that people will think about the possibility that at any time or any place they could be the next subject of the next episode. No criminal punishment to a person (except of course in extreme rare cases.) The central idea is simply that the people and/or thier vehicles will be recognized only by thier immediate center of influence. For the thousandth time, it is not my intention to snitch on people to line the pockets of law enforcement by providing them with easy means to fine and accost people. That is thier job to do, however ineffective that actually is. In fact, I think it would only have the complete opposite effect to what I am trying to accomplish.

I invite discussion of the ideas involved in this, and possible solutions to any of the obstacles it will inevitably encounter. However, I will simply dismiss any accusations of "...you are trying to be a snitch" because it would simply mean that you ignored everything I have said. If you have a rational argument to the contrary, I am open to hearing it. Perhaps, it will simply highlight a flaw in the logic that needs to be addressed to avoid it being used for that purpose.

Scott McDonnell

**RetroPlayer** · 04-10-2006, 04:45 AM

I should add that this is not necessarily an idea I am throwing out there that I want 'someone else' to do. I work on Flight Data Recorders as an Engineering/Technician for a living. I am great with hardware, but mediocre with SW (at least on a PC...assembly/machine language on micros-no problem.) I believe the SW involved is within my technical limits, and the hardware definitely is.

I had already thought of a name for the system (probably not original at all) which is 'iWitness.'

Finally, it is most important that the system is as anonymous as possible. This ensures the safety of the participants and subjects. Nobody should feel singled out or infringed upon enough to incite rage or legal proceedings. Of course, socio-politcal aspects of this are almost entirely dependent on the person publishing the results. Nothing you can do about that. However, the technology should take into consideration as much as possbile protections against the misuse -mistaken or otherwise.

**darkseid** · 04-10-2006, 08:55 AM

I have already designed most of what you are proposing. A VVR (Vehicle video recorder) with upto 16 camera inputs all capable of recording at 25/30fps (PAL/NTSC) and an audio channel per channel of video. Also there are at the moment 5 sensor inputs (12v input to interface directly with vehicle sensors) but the hardware is scalable upto 125 inputs. This kind of system in the UK is fitted to almost ALL buses and coaches to protect against anti-social behaviour and fraudulant insurance claims. The biggest problem for a system is to get the costs down to a managable figure for consumer level (in you scenario). At the moment systems like this inthe UK retial for around £1500 GBP which is about 2300 dollars.

**RetroPlayer** · 04-10-2006, 04:11 PM

That is awesome, but yeah, we have those systems for those same purposes here in the states as well. As you noted, though, cost is a bit prohibitive for this purpose. Also, the overall purpose of that system is meant to be 'big brother' and report incidents to the police.

I think the low cost is actually doable with sacrifices. Definitely doable if a carpc is the center of it (not sure I want to limit it to that, though.) The trick is coming up with the right balance of sacrifice.

I imagine only 4 cameras would be needed. Recording simultaneously might pose a bit of a problem. The FDRs that I work with obviously have separate channels for audio, video, and frequency related inputs. Discretes (similar to the 12V to TTL translation you are talking about, but at aircraft 28V) are muxed, though. The number of discrete inputs matters little actually. It's fairly cheap to add as many as you want (within reasonable limits of course.) The video portion itself will be the most challenging, I think.

How did you deal with that? Do all 16 cameras just record independently, or are they switched/genlocked/digital dsp?

The secret will probably be big buffers stacking up frames for processing, which means fast ram, which translates to moderately expensive. Question is - will this be cheaper than adding processing power?

Scott McDonnell

**TheLandYacht** · 04-10-2006, 05:04 PM

Originally Posted by RetroPlayer

You have a 5 lane road. Two lanes going in both directions and a center lane for turning left. This lane is shared by both directions of traffic. The concept is simple and safe if used properly. However, people abuse this lane when turning left into traffic. They pull into this lane to merge with traffic, sometimes driving down that lane for several hundred feet. This is EXTREMELY dangerous! If I could only convince you how many head on collisions are a result of this, and how common it actually is. I witness it (accidents of this type) almost once a week with my own eyes, and I am just a citizen driver.

Scott McDonnell

Your concept has merit. Your example of bad driving, however, doesn't hold true. In many states it is legal for a vehicle seeking to merge with traffic to enter the "fifth" lane of traffic, stop, and wait for a break in traffic to allow them to continue. Although driving down the center lane is patently stupid, the use of it as a merging tool is both sensible and legal in many areas.

In short, choose another example.

**RetroPlayer** · 04-10-2006, 05:18 PM

Originally Posted by TheLandYacht

Your concept has merit. Your example of bad driving, however, doesn't hold true. In many states it is legal for a vehicle seeking to merge with traffic to enter the "fifth" lane of traffic, stop, and wait for a break in traffic to allow them to continue. Although driving down the center lane is patently stupid, the use of it as a merging tool is both sensible and legal in many areas.

In short, choose another example.

Why? First of all it is illegal in more states than it is not. To be honest, I would really like to see where this is 'officially' legal..probably would be covered under wreckless driving if a cop bothered to enforce it, which they rarely bother to do even where it is clearly illegal. Just because a law is rarely enforced does not make it legal to break that law. Try arguing that in court!

Besides I specifically mentioned that driving down the center lane is very dangerous (and always illegal in some way) in my example, which you have eluded to yourself. I'm sorry, not to be a jerk, but YOUR example is flawed- not mine.

In short, it is a great example of a 'bad driving habit' which fits perfectly with the topic at hand (and doesn't HAVE to be illegal.) Whether it is illegal or not is really beside the point.

Scott McDonnell

**Enforcer** · 04-10-2006, 05:24 PM

sorry, what a load of cack.

**RetroPlayer** · 04-10-2006, 05:27 PM

Originally Posted by Enforcer

sorry, what a load of cack.

Thank you for adding a reasonable, educated, and well thought out response to a topic you could have clearly ignored if you really had nothing significant to add.

Scott McDonnell

**Enforcer** · 04-10-2006, 05:39 PM

Originally Posted by RetroPlayer

Thank you for adding a reasonable, educated, and well thought out response to a topic you could have clearly ignored if you really had nothing significant to add.

Scott McDonnell

well I could have made a long winded and puffed out description of why it wouldn't work, how many civil rights it would probably infringe, how impractical it would be and how your example probably means nothing to most of the world (it may be illegal in most states in the US, but not in most countries of the world). However I thought that the readers would have done enough reading and decided to give a short, sharp, and brief description of my opinion.

It would appear I have now had to extend it.

**RetroPlayer** · 04-10-2006, 05:52 PM

Yes that would have been welcomed, and I have no doubt you could have made some extremely reasonable arguments. However, you just simply assumed that you were right, and everyone else in the world, besides me, would just agree. In effect, trying to ostracize me. Again, the most effective tool of getting someone to self-regulate themselves. See if it were successful, I would simply concede and withdraw as I would be among my peers. Unfortunately, your method was just a little bit arrogant, really.

So please, unless you simply aren't interested enough in the topic, expand on your arguments. I don't mean that as confrontational. I really would like to hear them. I am sure if someone has 'read enough' they can decide that for themselves and simply click on the back button in thier browsers. I am almost certain they don't need you to 'save' them.

Scott McDonnell