Seeing Red?

In my past life I worked as an academic researcher, and I still tend to look at how to turn everyday questions into potential research projects. For example, take the problem of those who throw rubbish out of their vehicles. There is now a £150 fine for such behaviour, but the tricky part comes with catching people in the act and enforcing the fine. Could crowdsourcing be the way forward?

Imagine for a moment that CCTV is set up at traffic lights that are known to be littered. You don’t have to look far: most have a pile of coffee cups or food wrappers where the driver tossed them while waiting at a red light. The CCTV records a weekend’s worth of traffic, and the video is then cut into 5-minute segments for delivery to the computer of rubbish-watching volunteers. These members of the crowd watch the video and flag any time they see someone litter. When this occurs, the video clip is sent to the police, who handle tracking down the driver and issue the fine.

Think it sounds farfetched? This type of crowdsourcing is already taking place at Zooniverse, Mechanical Turk, and other sites. Indeed, you may not be surprised to hear that my own academic research was on this very subject—crowdsourcing to support environmental causes. Specifically I was looking at what motivated people to participate in such activities, and yes, some people are motivated by getting paid a token amount for their work. But many more enjoy competing with others or themselves, or simply feeling that what they are doing is worthwhile. Providing that the activity allows for basic gamification, such as a leaderboard or badges, it doesn’t take much to encourage participation.

There are, I admit, ethical considerations that must be addressed if a project like this is to be successful. The last thing I would want is for someone to play vigilante, splashing a litterer’s number plate across Facebook or trying to track down the perpetrator. But this could be mitigated by serving CCTV videos to those living on opposite ends of the country rather than within a local area. Indeed, it could even be international, with British volunteers watching streets in America and vice versa. I would hope that it wouldn’t take much of this—a year, maybe less—before word got out: littering is being taken seriously.

Some may grumble about the potential cost of this. After all, there’s the CCTV itself, the website to deliver video clips, recruiting participants, and perhaps a few people to oversee the programme as a whole. But what is the cost to the country—and the planet—if we don’t act?

[ Edited August 2019: I actually think catching people in the act of littering at red lights could be made even simpler by using the latest advancements in computer vision technology and machine learning. Any interested PhD students, please get in touch! ]

[ Edited April 2021: And it looks like LitterCam AI has successfully created this technology! ]

Leave a Reply

Your email address will not be published. Required fields are marked *