You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more
NB. most of my ideas have already been invented so this could be a very short thread.
There are now 1,000’s of cameras in cars, vans, bikes, commercial vehicles recording abortive footage waiting for an accident to happen. The reason all these hours are wasted is because nothing actually happens or in the case of a one off near miss or careless, bad, aggressive, vindictive, reckless, drunk driving no action will be taken by the under resourced police force. I am sure everyone has a moment of bad driving, however I am convinced that there will be a number of repeat offenders…… So
what if there was a web site where you could up load footage of bad driving which then grouped the clips by number plate, thereby linking the car to other incidents of bad driving. If the police were then presented with a dossier of video evidence of say 3 dangerous incidents by the same vehicle they could send letters out similar to speed camera fines. Some people will fess up, others may have their day in court and others will receive a tug from the police where they will undoubtedly discover other offences such as bald tyres, no insurance, skunk farms, slaves, stolen goods etc etc.
As well as having the ability to upload, users will also be able to watch uploaded footage. There are a number of reasons they would do this either through morbid fascination (Police Camera Action type footage) and also to see if they, or another driver, have starred in the latest production.
Are you still with me? Assuming so on to the technicality’s, I have no idea how to build such a system or whether it is possible to store hundreds of thousand of video files although these could be very short – restricted to no more than 2 minutes in length. One thing I grappled with for an age was number plate recognition however I have a very simple solution. The uploader just names their file with the offending number plate and then the clips can be easily grouped together. Even I can do this with the sort by tool in file manager so hopefully someone could write a programme to do this automatically. I also thought how to make the site appeal to those wanting to report bad cyclists - these could be uploaded with the location where the incident occurred, Police could then wait at the location at the usual time to catch a repeat offender.
Finally financials. I think there are a number of ways this project could finance itself and hopefully make me some money. My ideal solution would be to take a load of positive responses from STW to GoPro, Garmin, FlyQ, Google and they compete to buy the idea and make me a millionaire overnight, however I think this is unlikely. My second is for STW to come to the rescue and I finance someone to build a cheap basic version of the site which we launch and sponsors pay to advertise on the site - if it takes off then my partner and I sell out. The third route is that the insurance companies take an interest, similar to Avia’s safe driving app, and see the web site as a preventative measure that would reduce future insurance claims.
What do you think ? I am happy to be flamed and put the idea out of my head. Alternatively if there is genuine interest I am able to provide a modest amount of finance to pay for a partner to try and help me move forward on the web production side. If you want to nick the idea I will obviously be disappointed but at least something may come to fruition and help rid the road of bad drivers.
Technically easy. YouTube plus something like Azure video analytics plus Logic Apps to chuck the number plates and date time info in a data store.
The rest... There is already a "near miss" logging site, have a look at that.
Not sure on whether you'd be able to drive any legal action off it though.
I've heard many many worse ideas.
There is bound to be some legal tape surrounding Data Protection. I mean, you're storing personal information about someone without their knowledge or consent. Information will end up in the public domain. At a minimum we've location and license plate. Over the course of a few mins of video you have a rough idea of the direction they're heading in. There's the risk you'll have someone's home if the incident occurred nearby to it. Then there's the information you can get based off the license plate itself. Which for a mere £10 is the full service and tax history. I've not tried to do much by snooping but I'm sure there's much much more out there.
So the question becomes, if it's needed, how is it kept in a secure manner? Well, the short answer is, it isn't. Nothing is secure. So you're looking to set up a system that, if abused, can put a lot of people at risk. And we've not considered how it could be abused. Don't get me wrong, it's an awesome idea to actively remove dangerous driving... but nope.
There is bound to be some legal tape surrounding Data Protection. I mean, you're storing personal information about someone without their knowledge or consent. Information will end up in the public domain. At a minimum we've location and license plate. Over the course of a few mins of video you have a rough idea of the direction they're heading in. There's the risk you'll have someone's home if the incident occurred nearby to it. Then there's the information you can get based off the license plate itself. Which for a mere £10 is the full service and tax history. I've not tried to do much by snooping but I'm sure there's much much more out there.
So the question becomes, if it's needed, how is it kept in a secure manner? Well, the short answer is, it isn't. Nothing is secure. So you're looking to set up a system that, if abused, can put a lot of people at risk. And we've not considered how it could be abused. Don't get me wrong, it's an awesome idea to actively remove dangerous driving... but nope.
Really! I don't think a number plate displayed on a vehicle for ALL to see can really be considered personal data.
[url= http://rate-driver.co.uk/ ]Here? - it's even sweary, so got to be good[/url]
Then why won't Google Earth/Maps allow you see license plates? They obscure them, along with faces. It's personal data, what's more it's belonging somebody else. What gives you the right to let the world know that a gold BMW was turning onto the M whatever and cut you off? That car will be identifiable, and thus the information could be used to endanger the life of the driver. This is why there are data protection laws. You are only allowed store certain information and in a very specific manner, for a limited amount of time. None of it can be stored without consent.
[edit] Well, that's me told. I'm still not too happy about it as a concept.
Figure it's treading a fine line in grey areas legally.[/edit]
Really! I don't think a number plate displayed on a vehicle for ALL to see can really be considered personal data.
The difficulty, I suppose, is the numberplate identifies the car, not the driver. If you group a set of clips together to imply a dossier of damning evidence (because they all feature the same numberplate) you don't know if they're all the same driver or if you are tarring other perfectly well behaved drivers of that vehicle with the same brush
In terms of a business that could be sold on or used as an income stream from advertising you're going to be looking at getting volume users. How many people are likely to upload their dash cam footage when they encounter 'bad driving', what percentage of the population use dash cam anyway, is it high?
Also in terms of the compiling dossier thing how likely is it that three people encountering bad driving from 'car a' will be recording it, will upload it to the internet and be registered on your site.
Slim I would say, and because of that I'm out.
Also how could the police act when there is no evidence that it's the same driver in each sequence of footage. And how would they find time to review the footage and act on it?
Then why won't Google Earth/Maps allow you see license plates? They obscure them, along with faces. It's personal data, what's more it's belonging somebody else. What gives you the right to let the world know that a gold BMW was turning onto the M whatever and cut you off? That car will be identifiable, and thus the information could be used to endanger the life of the driver. This is why there are data protection laws. You are only allowed store certain information and in a very specific manner, for a limited amount of time. None of it can be stored without consent.
[edit] Well, that's me told. I'm still not too happy about it as a concept.
Figure it's treading a fine line in grey areas legally.[/edit]
I still disagree.
1. Can you identify the individual from the number plate? No.
2. All cars are displaying this info to the public at all times anyway.
3. I could walk down your street, see you car parked outside your house (I know where you live) and search your car reg online. So what?
Biggest problem is that now that the internet is so damn big, pooling the resources into one Internet site just wouldn't be possible. You'd have to pay Google to have it as the no. 1 hit, or advertised on the search results page.
I'm sure there are already sites attempting to do something similar (howsmydriving.com etc), as well as Facebook pages, YouTube channels etc.
How do you intend to be the no.1 site for bad driving videos?
You haven't planned this?
For that reason.. etc.
No problem I can see from Data Protection - it's totally legal to record videos of anything happening in a public place* and totally legal to publish such videos. Number plates don't count as personal information within the scope of the DPA.
Google Earth is a bad analogy - their policy is probably based on US law if it's based on any law at all and not just their decision. There's not even any reason in this country why you can't upload pictures of people's faces (you might notice there are plenty of those on the web already).
* with some limited exceptions
Here? - it's even sweary, so got to be good
First STWer to find themselves wins!
Anyone know Ronnie Pickering's reg?
Who?
Here's my better idea for free and disruptive
Allow blanket upload of all camera footage with AI to find and report the bad driving. Should not be too hard to do via machine learning. Give the driver a score and rating that they can share via social network or with insurance company.
As a result create an alternative to Google streetview, a pothole scanner, suggestions to HA for road improvement etc...
Please name your first yacht after me.
Alright - think about how you can do this in a LEAN way. Chances are you won't even have to build anything to get a prototype up and running.
1: read "The Lean Startup" by Eric Reis.
2: design your product using the least resource possible.
I would look into maybe hosting a YouTube channel where all videos are named and tagged in a specific order so that it's easily searchable for multiple instances of the same licence plate. You then ask on forums/facebook groups for people to upload their videos following your naming format and you can edit/moderate the content
(As no doubt some will screw up the naming or upload unrelated content - worth noting this will happen with any solution, so it's worth figuring out the human resource needed here. You might be able to build or licence some image recognition software to automatically detect licence plates but lets assume you have no money.)
Doing this will validate a number of things:
1: People will be willing to upload their videos. If your site doesn't have users uploading videos, you don't have a product. If people won't upload to a YouTube channel/group, there's an even smaller chance they'll upload to another service that isn't as big/truthworthy/known/etc as YouTube.
2: Any potential privacy concerns. YouTube will quickly be in contact if they think you're violating others privacy. You can use this information to go forward.
3: Desire from law enforcement to act on anything. Great - you've got 3 videos of the same guy being a dick - does the Plod give a shit?
4: The frequency that someone gets caught more than once. It sounds plausible, but I'd like to know for certain how often a video of a previously filmed driver will be uploaded before I invest any of my hard earned cash.
If these things validate in a positive way, then and only then would I consider putting money into building a product.
And finally. Advertising is [b]not[/b] a revenue model. Do you have any idea how many thousands of impressions you need to make any substantial revenue from advertising? There is a reason why all news publications have started putting up paywalls (because advertising doesn't work)
Sponsorship might work - but then you need someone to go round and rally up willing sponsors and to make sure there is always money coming in - which is another salary on top of your developers.
As a sustainable business, I'm out. You might put together a pitch and try to get a grant from local councils perhaps but I don't think this will make you any money. The companies you listed have no reason to buy you out, your exit strategy is wrong i'm afraid.
Could this not be done on twitter. Use the number plate as a hashtag.
very easily searchable
That's also a great idea.
#crapdriversUK #<numberplate> #<localtown> @<localPlod>
I'm out.
IME, the WORST drivers are those with the cameras in their cars.....
(typically of the "look, i'm clearly stupid enough to be (deliberately) driving up the inside blind spot of this huge truck, oh, what a surprise, he's cut me up, the driver of that truck is a T**T" kind )
Too many barriers to gathering the volume and accuracy of the data you need to make it a viable resource, let alone the concerns around privacy and abuse. What is the driver for people to contribute if there is no guarantee of action?
What's more, it could be possible in some circumstances for the filmer / contributor to be identified by the accused, for example if they are regulars on a particular route or the accused has a camera themselves. Not a great incentive.
Does this not already in effect exist?
Every other helmet/dash cam warrior seems to tag the videos of their adventures with date, vehicle make and Reg No...
TBH though if you manage to get footage of a motoring offence it should really be forwarded to the relevant police force... Right?
Dragons
Thank you for your feedback, I really appreciate the time you have taken to review my proposal and give your considered opinions.
I believe my idea is sinking but not quite sunk and would provide the following responses for your further thought. In order of holes below the waterline;
Is there actually enough bad driving for 3 individuals to record and upload 3 different incidents. I agree if this is true this is would be a slam dunk sinker. From reading threads on hear you would believe there were incidents every day, from my own experience of commuting an average of 100 miles a week into London – I would say on average 1 incident involving bad driving per week (10’s if you include cycles but that is another debate). I really like plypon’s LEAN suggestion but even this is beyond my IT skills would anyone be prepared to help me (for a modest fee) set up a YouTube channel to host videos assuming I can resolve the points below?
DATA Protection – never even occurred to me. Worthy of a note from lawyer to confirm one way or another.
Revenue. I thought advertisers paid per click on the website ? Without revenue I am out!
Competition/does the idea exist - I couldn’t find Near Miss; rate-driver.co.uk quite useful and you could search by Registration plate – but does not feature video or grouping of incidents so cannot be used as evidence; Howsmydriving.com nothing similar came up, I therefore agree it would need potentially expensive promotion to get to the top of Google – I need to research the costs.
Volume of Data – I thought this might be a problem. Does anyone know how this could be quantified?
Would the police act on a dossier of examples of bad driving – surely it would be hard not too? It is such a cost effective way of patrolling the roads. 30 mins to review footage and issue a letter. If not could the insurance companies or individuals pursue a private prosecution?
Identify the car not the driver. This is no different to speed cameras where there is a presumption the registered keeper is the driver and they have to prove otherwise or face perverting the course of justice.
Thanks again for your thoughts and if anyone could help bottom out the outstanding items this would be much appreciated.
Ben
How are you going to Police it? Before you know it you will be having to watch crappy videos day in day out.
Ie how are you going to stop people uploading porn vids etc
Also how do you stop selective posting of vid. e.g. Video person nearly drives other car off road but doesn't post that bit of the video.
Who sets the criteria fir what is considered dangerous?
Daft idea