Lichtvisie: A Look at Amsterdam’s Public Lighting as an Urban Sensing Platform
Four months ago, I went to Amsterdam with fellow students from MIT and Harvard Graduate School of Design for a spring workshop (11.320) taught by researchers from the Senseable City Lab. The aim was to explore urban lighting infrastructure as a potential “smart” urban sensing platform.
I can be a bit of a nervous flyer, especially on evening flights, but one thing that does bring me ease is catching a glimpse of a city from above.
The sight of city lights signal an arrival. Human presence. Or somehow — life itself.
Public light, in many ways, is meant to affect comfort and relief, for its connection to safety is as old as the lighting of public space itself. During the Middle Ages in Europe, people were required by law to carry lanterns after curfew to prove they were not committing any crimes or lewd acts in the late of the evening (Seitinger 2010). You carried a light to let others know that it was safe, that you weren’t dangerous.
“Anyone who is found at an unusual hour in an unusual place without a light must submit to the strictest investigation.” (Lehrbuch der Staats-Polizey-Wissenschaft, Johann Heinrich Jung, 1788) (Schivelbusch, 1988, p. 82)
A pervasive fear of what lurks in the dark also persisted in the public imagination. The duality of light as inherently good and dark as inherently bad is best depicted by an image printed in a 1910 edition of Electrical Review and Western Electrician. In it, a devilish figure, accompanied by other archetypal criminal characters, squints (and almost hisses) at the sight of a public street light and turns away from the glow. The anthropomorphic street lights take on the shape of policemen — a clever visual metaphor for how public lighting is perceived to eliminate chicanery, violence, and death lurking in the darkness.
Lights eventually moved from being carried on our bodies to being embedded in our surrounding infrastructure, first as gas lamps lit by kerosene and then as bulbs powered by electricity. Whereas public spaces previously became vacant after sunset, public light’s continued illumination of urban space after twilight enabled longer hours of socializing, commerce, and transportation. In effect, the history of public lighting runs in parallel — pun intended — with the history of the modern city.
In present day, public lighting is one of Amsterdam Smart City’s major areas of investigation. Following one trend toward sustainable energy, a number of experiments in urban design have looked at how to make public lighting more efficient. For instance, many cities like Boston have taken to replacing old sodium street light bulbs with LEDs, which last much longer and consume less energy. Equally popular are “smart grid” systems that allow for monitoring of energy usage and even remote programmability of city lights. In Amsterdam, an “on-demand” lighting system called Tvilight designed by Delft University of Technology is an installation of public light fixtures in which the lights only turn on in the presence of a person, bicycle, or car.
But what also makes public light interesting is that it is a ubiquitous form of infrastructure. Its placement largely depends on the flow of people traffic. With the advent of pervasive sensing technology, alongside a desire and need for the real estate on which to deploy various sensor nodes, public light might be an ideal platform on which to deploy such a network. Other “smart city” projects like Chicago’s Array of Things have also conceptualized lighting infrastructure as a sensing platform.
This is what became the core research question for me while we were in the Netherlands:
For which contexts could public lighting-as-sensing-platform be appropriate for Amsterdam?
Off the bat, I will hedge and say that our time in the city was limited. We were in Amsterdam for all of four days and jetlagged for one of them. By no means did we perform extensive ethnographies by immersing ourselves in the environment longitudinally. However, we did what we could for the limited time that we had for this workshop assignment.
We met with urban planners, representatives from Amsterdam Smart City, Philips Lighting Company, and the Citizen Data Lab at the University of Applied Sciences. On our own, we conducted street interviews and participant observation of spaces with urban lighting infrastructure and made contact with people who might be willing to keep in touch after we were back in the States.
One common pain point among city planners and citizens alike was the issue of overcrowding. In some conversations, this was referred to as the “Disneyfication” or “Rome-ification” of Amsterdam, meant to convey that the city was simply becoming a place for tourists and not for residents. Many locals I spoke to reported seeking alternative commuting routes to avoid crowds of tourists altogether.
While exploring the city, we had many non-automotive options for getting around. We, of course, hired bicycles to get around for a couple days, but we also spent a lot of time on foot and on public transportation. A few times, I noticed pedestrians with crutches and during one instance, I noticed a pedestrian who was visually impaired and needed to use a cane to navigate the city.
This led me toward thinking about accessibility in cities, and whether or not “smart cities” accounted for accessibility.
Was designing for access an afterthought in the smart city dynamic, or could it be something one led with when thinking about integrating technology and urban space?
While this was certainly of interest to me from a research perspective, with an urban design cap on, I needed to further investigate whether it was of interest to others in Amsterdam. I started digging deeper, and I started talking to people about it.
In the Netherlands, there are 311,000 people with vision impairment, approximately 2% of the entire population. There are also an estimated 30 million visually impaired people in all of Europe.
In the end, I had ten interviews with (A) people who identified as visually impaired and lived in Amsterdam and in other urban settings and(B) people who have worked with visually impaired individuals to train them how to navigate urban spaces. Some common things I heard were:
“I’ve used BlindSquare for the iPhone. It was helpful but didn’t give me information in real time. It also took a long time to use because of Voiceover.”
“In almost all cases, visually impaired pedestrians preferred not to be assisted, with the exception being when entering an unfamiliar place.”
“My GPS and headphones work well when I’m lost.”
The cognitive load that visually impaired pedestrians must pay acute attention to — street names, where the street is in relation to the sidewalk, number of lanes in the street, direction of traffic, when to cross the street, whether they are on the left or right sidewalk, nearby buildings and traffic light, etc. — is further compounded by having to navigate crowds of people, on top of also having to navigate nonhuman obstacles.
For the purposes of the workshop assignment, I started to focus my design research on how urban lighting as a sensing platform might be generative for urban accessibility for visually impaired pedestrians. What would a speculative design for such a system look like?
The iteration process had begun.
Having heard from interviewees that audio was their preferred method of receiving information to navigate, the first design I drafted used light poles to sense and convey peripheral information to visually impaired pedestrians. A pedestrian might then be able to hear a message about what was directly around that pole — the name of the street they were on, the intersection name, or simply whether there is traffic up ahead. This could make it easier to get context about unfamiliar spaces.
A subsequent version adjusted the system so that visually impaired pedestrians could “tap” light poles when they were lost and receive similar navigational information. However, for both versions, the initial rounds of feedback from my peers and some interviewees revealed this might not be practical, as many visually impaired pedestrians are quite methodical about navigating spaces and veering off-course to listen to directions might be counterintuitive.
Finally, after much more research, much more feedback, and likely too much coffee, I arrived at the final proposed design.
The final “Verlichten” proposal involves using 360 cameras deployed across the network of lighting infrastructure in order to monitor pedestrian, cyclist, and vehicle traffic on the sidewalk and road. Because street lights are fairly ubiquitous and distributed along pedestrian spaces, they can be leveraged as locative sensor nodes that help orient visually-impaired pedestrians. Each light would be “aware” of its vicinity and be able to relay that information to individuals with the companion mobile app.
Because visually impaired pedestrians usually cannot see the screens of their smart phone screens, auditory and verbal features of smartphones like Apple’s Voiceover help with navigating the interface. The mobile platform would be designed to “speak” to the visually impaired pedestrian in real time using text-to-speech technology as s/he proceeds along the commute, communicating information about oncoming pedestrian, cycling, or vehicle traffic. Pedestrians would be able to access this audio via the smartphone’s speaker or a headset.
The sensing platform would be aware of the pedestrian’s location at all times as s/he enters the sensing radius of each light pole and communicate information about the pedestrian’s immediate vicinity.
Each pole’s camera would locally sense its surroundings for people and vehicle traffic; extract count data from low-resolution images; and relay the information to a mobile platform that can be used by visually impaired pedestrians. In order to account for privacy, cameras would be low-resolution for purposes of protecting privacy, and no images would be archived. As soon as images are captured, they would be analyzed, then discarded. Count information extracted from them would then be passed on through to the server, without sending the image itself.
The site for the design would be the Knowledge Mile, an area in Amsterdam slated for development into an innovation district by 2018. It spans the length of Wibautstraat on the southeastern side of the city. The Knowledge Mile would be considered a “simple” structured environment (Gaunet 339) wherein crosswalks are perpendicular and sidewalks are parallel to the street. Unlike downtown Amsterdam, the Knowledge Mile has wider sidewalks and can take a longer amount of time for pedestrians to walk from one block to the next. Most visually impaired people live and travel in such environments, making it an ideal test site.
Verlichten in Dutch means “to illuminate,” which can be read two different ways — both to irradiate a surface or to make something known and bring forth knowledge. For this assignment, I came up with a speculative design that uses lighting as a sensor platform to act as eyes on the street for those who cannot see. (Once again, this was done after spending only a limited time in the city but quite a bit of time doing interviews thereafter with people I had contacted while “in the field.”)
As cities forge a path into a digitally-oriented future, a “design for all” approach in addressing overcrowding in the city might ensure that this future would be one accessible to underserved populations as well as the general populi.
“Leaning on a Lamp-Post” (1937)
performed by George Formby
I’m leaning on a lamp, maybe you think, I look a tramp,
Or you may think I’m hanging ‘round to steal a motor-car.
But no I’m not a crook, And if you think, that’s what I look,
I’ll tell you why I’m here, And what my motives are.
I’m leaning on a lamp-post at the corner of the street,
In case a certain little lady comes by.
Oh me, oh my, I hope the little lady comes by.