Interactive Lightstreet

1. Overview

Interactive Lightstreet is a modular, sensor-driven LED lighting system embedded in the pavement of the ArenA Boulevard in Amsterdam, the stretch running past the Amsterdam ArenA, Ziggo Dome, AFAS Live and IKEA in the Bijlmer/Amstel III district. Rather than lighting the street uniformly from overhead lanterns, the installation detects pedestrians as they walk and illuminates the ground immediately around them, creating a moving pool of light that follows each person along the boulevard.

It was commissioned by Co-ReUs, a research and co-creation initiative, in collaboration with the University of Applied Sciences of Amsterdam (HvA) and the Amsterdam municipality, as part of a minor in Intelligent Environments within HvA’s Technical Informatics programme. I built it together with Mathijs Haakman and Mitchell Ebbers. There is no public source repository for this project; this document is based on the original design research write-up, published as a portfolio article.

2. Problem & context

The ArenA Boulevard scored 122 on Amsterdam’s insecurity index against a city average of 108. Long observation sessions from the roof of the nearby Deutsche Bank building showed pedestrians consistently walking far apart from one another on the boulevard’s wide, open pavement, a pattern that environmental psychology research on public-space lighting links directly to feelings of vulnerability [@nasarbokharaei2016]: walking alone through a low-quality public space reinforces the sense that the space is unsafe, and vice versa. That line of research sits alongside a broader body of work on street lighting and perceived safety, including Welsh and Farrington’s review of lighting improvements and crime [@welshfarrington2008], Painter’s studies on street lighting and pedestrian use [@painter1994], and Flynn’s early work connecting lighting to perceived safety in public space [@flynn1988].

Conventional overhead street lanterns light the boulevard in high-contrast pools separated by dark gaps, rather than offering pedestrians a continuous, dynamic field of view onto the people around them. The working hypothesis for this project was that lighting the ground where people actually walk, rather than a fixed grid of overhead pools, would draw pedestrians closer together and measurably improve their sense of safety.

3. Concept & interaction design

The system runs in two modes:

• Idle mode. After 1.7 seconds without a detected presence, the strip plays a randomizedDots animation: thirty random LEDs per rail light up in random colours for two seconds each, meant to draw pedestrians toward the light street from elsewhere on the boulevard.
• Interaction mode. Ultrasonic sensors detect a pedestrian entering an active zone between 20 cm and 220 cm (with a maximum sensing range of 701 cm) and trigger the LEDs around their position, following them as they move through it.

The interaction design was deliberately built around Albert Borgmann’s Device Paradigm [@borgmann1984]: the goal was a transparent black box, where a pedestrian could intuitively grasp the cause-and-effect link between their own movement and the light’s response, without ever needing it explained. That legibility was treated as the main lever for user satisfaction: a light street nobody understands is just a light street nobody trusts.

4. Hardware & system architecture

The installation used four two-metre LED strips (60 individually addressable LEDs per metre, 240 LEDs per rail, roughly 8 m² of lit surface), each rail driven by an Arduino. Battery-powered, 3D-printed sensor cases, not weatherproofed, this was a field prototype, not a permanent fixture, housed the ultrasonic rangefinders and polled at 10 Hz over XBee wireless modules, aggregated by Sodaq boards before reaching the Arduinos driving each LED rail.

flowchart LR
    ps(["Pedestrian in active zone<br/>20-220 cm"]) --> us(["Ultrasonic sensor<br/>battery-powered, 3D-printed case"])
    us -->|"XBee · 10 Hz"| sd(["Sodaq aggregator board"])
    sd --> ar(["Arduino per LED rail"])
    ar -->|"no presence · 1.7 s"| idle(["Idle mode<br/>randomizedDots animation"])
    ar -->|"presence detected"| lit(["Interaction mode<br/>light follows pedestrian"])
    idle & lit --> led(["4x 2 m addressable LED strip<br/>60 LEDs/m · 240 LEDs/rail"])

5. Field experiments

The system was tested outdoors on the ArenA Boulevard itself at dusk (10 p.m. to midnight, on weekdays, roughly 20°C), near a boulevard bench that offered pedestrians a natural binary choice between two parallel paths:

The deviation study, in particular, treated the underlying hypothesis carefully rather than declaring victory: the light street clearly changed how people moved through the space, which was the measurable proxy for the harder-to-measure claim that it made them feel safer.

6. Presentation

Interactive Lightstreet was installed and tested in situ on the ArenA Boulevard rather than shown in a gallery; the boulevard itself, at night, with real pedestrians, was the exhibition context. The full design research, including the rooftop observation study, the walking-line analysis of 165 respondents, and the three field experiments above, is written up in the original article, “The Influence of Light in the ArenA Boulevard”.

7. Reflection

The honest conclusion drawn at the time still holds: the light street offers a useful tool for altering walking lines and can be used for future experiments and research. Not every pedestrian read the interaction correctly on first encounter, which is exactly the kind of signal a transparent-black-box design is supposed to surface. The failures were as informative as the 64.38% who were measurably drawn toward each other by the light.