Design for Recognition explores how image recognition technologies shape the ethics of being seen.
When cameras with image recognition are deployed, for example, in parking scan cars or scan bikes, this often makes a process more efficient. But specific data or statistics about these savings are not available to residents or visitors. Why not? Could radical openness also advance the debate about the use of this technology and lead to new insights among the public?
Now that parking enforcement is being carried out by a parking scanner car instead of an officer on the street, there's no one around to ask questions. The officer is currently in the office, at the computer. In between, they have to wait long periods until new photos of a potential illegal parker appear. Can the officer answer questions at the same time? Could we set up a helpdesk for people with urgent questions?
People often think they're being filmed when they see cameras. Sometimes they wave at them. Usually, photos are taken as soon as an object is detected. So, no video clips are saved, and no one is watching live. To make this clear, you can make it visible when a photo is being taken. For example, with a light or an opening aperture. It's important that the "signal" is clearly visible in daylight, or in the case of a parking sensor, on a passing object.
Street cameras often point in one direction, but that's not the case on a parking scanner car. They're even somewhat hidden. The camera box on the car thus becomes a kind of "all-seeing eye"—but of course, that's not the case. Could we better communicate its operation by adjusting the shape of the camera box?
Various tools and discussion formats are available for public consultation sessions about physical changes in the city. Consider, for example, a scale model that everyone can walk around to assess the plan's impact. Such formats don't exist for digital changes in the city, even though these types of plans can still impact public spaces. Can you create a digital scale model for participatory meetings about (invisible) technology in the city?
Municipal technology projects are often only released to the public once they're completely finished. That's understandable; you don't want to make a blunder. But there's also value in presenting an unfinished product to the public. By presenting a mock-up as early as possible, you can learn how people perceive the product and what they think it does. You can then adjust the design accordingly. A conversation on the street can also offer new perspectives: for example, does it also work well for the elderly? Or for people in wheelchairs?
Coalition agreements, policy documents, and implementation agendas often outline ambitions and public values applicable to the use of technology in the city. During a participatory meeting, you can ask residents for feedback. It would be helpful if residents had clearly identified these democratically agreed-upon values, so it's clear how the proposal should be assessed. The value cards are a useful tool for this.
In a digital system like the parking scanner car, every exception must be coded. For some exceptions, this isn't a problem because they're common. Consider, for example, people who park and pay five minutes later. Such an exception is built into the algorithm. But there are also exceptional situations where you receive a fine, even though most people actually think it's unnecessary. In these situations, you could say that the human touch is missing. We organized a quiz where a group of people voted on the justification of a fine. Could such a human feedback loop also become part of the digitalized process?