About a year ago we got involved in a project called Telraam. According to the website, it’s a citizen-powered solution for collecting multi-modal traffic data with a purpose-built, affordable, and user-friendly device. OK, so that was a bit of a mouthful! In short, it’s a camera and a mini computer. It’s Dutch for Abacus and you’ll see why.
The device can count passing vehicles, categorise them by type (car, truck, two-wheeler, pedestrian) and even distinguish speed and direction of travel. It is mounted on the window of your home pointing towards the street, then algorithms process images from the outside world and generate data reports. We had a look at the map of sensors across the world and most are in Northern Europe and North America - perhaps not surprising as it is an EU-funded initiative. In London there seems to be a hot spot of cameras around Tulse Hill. A nice thing about the programme is that anyone can see data from any sensor - this is ours.
What about (UK) GDPR?
In his day job, Sacha is heavily involved in data privacy and information security, and one of his immediate questions was around data protection. If we have a camera pointed at the street, what are the data privacy implications? Telraam has a detailed explanation of it, but in brief, the unit processes low-resolution images in which licence plates and faces cannot be identified. These images are immediately categorised so no personally identifiable information is ever captured or stored.
So what insights did we glean? This chart shows the data from our sensor for the past two weeks. We noticed that the number of two-wheelers is far higher than the number of cars - 4807 vs 2563 (note that the number of pedestrians is inaccurate because of an obscured view of the nearside pavement). There is also a function where you can compare to other locations - this shows our car count is lower than average and our two-wheeler count is about average. This must be good for air quality - even if the two-wheelers are motorbikes, they are lighter than cars.
Another intriguing point is the speed of vehicles: the v85 speed (the speed below which 85% of vehicles travel) is 25.7mph. The speed limit on our street is 20mph, suggesting that it is not very well adhered to. This matters for air pollution because a lower speed limit encourages smoother driving with less acceleration and braking. While a recent study found no net reduction in exhaust emissions in a 20mph zone, it did find that this smoother driving style reduces particulate emissions from tyre and brake wear - which still represents a significant cause of air pollution from zero-emission vehicles.
Counting on you
Could all London neighbourhoods be as Telraam-ed up as Tulse Hill or even rival Benelux? We think so. There are a couple of ways to get started: there might be a network near you that is already counting which you can join; or you can buy a V2 device yourself for about €200. Scroll to the bottom of this page to see the options. If you’re technically minded, you can also build a V1 device yourself (which is what we did). It’s more fiddly to get going and doesn’t look quite as nice, but works just as well and is a lot cheaper - especially if you have a spare Raspberry Pi in a cupboard somewhere.
Having mounted this on our window, we now just need to get the air quality node mounted on a Brixton lamppost. The node is charged up, the ladder is ready... photos to follow!
Amazing! What a useful device. I’m sure some of the other groups would be interested! 🚗📊