Case Study: Traffic Lights and Smartphone Sights

Ground Level Traffic Lights
Ground level traffic lights, taken from

In which Hannah and Ben discuss the merits of reflecting user behaviours rather than resisting them

HK: At least three towns that I’ve heard of in the last year – in Australia, Germany and The Netherlands – have embedded traffic lights at pavement level to warn Smartphone users that they are passing onto a road.

Of course, I found this on the internet…  so groups are either praising it as a great idea, or they’re vehemently arguing that solutions like this are just spending a lot of money to help ‘bad behaviour’. There is a tendency to throw moralities at new technologies, whether they are inherently harmful or not, but it’s true that smartphone users don’t always have their sights set upwards, at street signs and corners, or road edges.

BK: One of the funnier sights is seeing a smartphone user walk straight into a bus shelter or lamppost.  Except when it was me!

HK: What a divvy. [Trans: ‘Idiot’ in Geordie dialect]. Though a lot of people I’ve met (including me sometimes) don’t know the names of streets they pass through daily because they’re using lines on their phone map and GPS to guide them. One article, describing the initiative in the Netherlands, calls them ‘smartphone zombies’, which comes across as inherently pejorative. But why shouldn’t cities reflect that actual habits of their citizens?

BK: Well, I guess that you could work that two ways – to what extent should the cities adapt to the habits of the citizen, versus to what extent should the city shape the habits of the citizen.  In this particular area – the very existence of kerbs and traffic lights shape the habits of citizens, in order to separate pedestrian and wheeled traffic. The ‘woonerf’ areas in the Netherlands, shared space in the UK or shared zones in Australia and New Zealand form a contrast. Isn’t this traffic light discussion another version of experimenting with adapt to, vs attempt to shape (aka nudge) vs attempt to control? I put three options in deliberately – as if you have two in, then you are limited to “either or” – a dualist approach which limits our thinking.  About as stable as a two-legged stool.

HK: Back to the subject please.

BK: Ah well. Anyhow whilst there are all sorts of positive outcomes of adapting to citizen habits, even with a bit of nudge: we have to take secondary consequences into consideration. For instance, the £1M cost to the NHS of falls because of the new ‘healthier/low pollution’ trams in Edinburgh; the public health benefits around encouraging more physical activity, vs ‘keeping people safe’ with kids walking to school.

HK: Then again, floor level traffic lights might reduce car/pedestrian collisions, safe police and medical services time and money. Inserting safety measures into residents’ lines of sight does seem pretty standard practice, like making sure stop signs can be seen from car level, or traffic lights at side-window level for the front car in a queue. – a habit in France, but not in the UK. Attempts to change the pattern to cope with ‘smartphone zombies’ could, I suppose, use other means,like more intense ‘crossing’ sounds, for example.

BK: But these could be could be startling and harmful to anyone else in the area. Light change seems fairly innocuous. Though – consider the partially sighted!  Road markings, and traffic lights, should be visible in all contingencies.

HK: And implementation for embedded lights is quite dear. There must be better options than digging them into the road. For example, there’s a huge trend right now into upscaling city lighting systems with new nodes, effective sensors [which should be next week’s blog] – why not attach a projection of floor level lighting?

BK: A colleague of mine – Steve Smith at Northumberland – pointed this one out – if you project from existing street furniture, then you don’t have to dig into the road surfaces; you can project from a point so it’s much easier to maintain; and the effect works (though admittedly not as effectively) even if the surface is covered with mud, water or snow.

But that’s just the practical. We have to consider all citizens – from the harried parent with children in hand, the hearing or sighting impaired, through to the inattentive mobile phone user. If we were being really sophisticated, we could consider combining sensors with warnings – so if someone stepped across a line into a dangerous junction, then not only would the lights under their feet flash, but a warning sound could go off.  

HK: So it’s passive and quiet – until needed, though I’d imagine that software like that isn’t cheap. But we’ve established that our cities aren’t full of zombies, with strategists as the brains:

BK: Well, sometimes around the Bigg Market in Newcastle on a Saturday morning?

HK: I’ll give you that. But civic action is most effective at keeping its citizens happy and healthy if it responds effectively to all behaviours, and their dangers.

BK: Seriously, a major mistake – at least in my opinion – is to think of a city, or indeed, any sufficiently complex system, as a machine that can be designed in which each cog behaves. It isn’t.  It’s an active mix of behaviours, interactions and decisions. This is an evolution of behaviour problem, not an engineering problem.

Setting up an environment in which that independence of decisions is understood and respected, but people are firmly guided away from behaviour which damages others (in this case, speeding, jumping red lights and jaywalking without care) requires us to consider each decision maker and shape the communal environment such that we strike a balance we can all live with.  

HK: So the advantage of ‘Smart’ in this case is that we can adapt that environment more accurately to a greater variety of behaviours?

BK: Yup.  So we can support a richer, more diverse environment.  Which loops neatly back to our second blog – ‘Why bother with Smart’’!