Чтение и письмо: аналитическая статья и предложение (модальные ограничения)

WHY SMART CITY SYSTEMS DISAPPOINT — AND WHAT THE SUCCESSFUL ONES HAVE IN COMMON
Over the past decade dozens of cities have installed «smart» systems: sensors that time traffic lights, cameras that spot illegal parking, software that predicts where bins will overflow. A review of forty such projects, published this spring, found that fewer than a third were still running as intended five years later. The failures had less to do with the technology than with the conditions under which it was introduced.
The first pattern is a benefit measured in isolation. Adaptive traffic lights in one city cut journey times on the main road by 12 %; the knock-on effect was that drivers abandoned the ring road, and congestion moved into residential streets, where nobody had installed sensors. The gain was real, but it was measured only where the system was looking. Successful projects, by contrast, defined the side effects they would monitor before switching on — and several were switched off again when those effects appeared.
The second pattern is consent that was never really given. Parking cameras in another city were introduced «for enforcement» and were later used, without public discussion, to track the movements of individual vehicles. Legally the council was within its rights; politically the scheme collapsed within a year. Where systems survived, their purpose had been written down in advance and could not be extended without a public vote — a sunset clause, in effect, on every new use.
The third pattern concerns what happens when the system is wrong. A bin-prediction tool sent crews to empty half-full bins while full ones on unpredicted streets overflowed for a week; residents could not reach a human who was permitted to override the schedule. The projects that lasted all shared one feature: a named person who could switch the system off, was required to review its errors monthly, and was accountable — in the ordinary sense of having to answer questions in public — for the decision to keep it running.
None of this is an argument against the technology. The review's authors conclude that the same systems, introduced under the right conditions, delivered most of what was promised. But «the right conditions» turned out to mean something specific: side effects named in advance, purposes fixed in writing, and a human who could say no. Cities that treated these as optional extras rather than mandatory design features were, five years on, the ones with expensive sensors and nothing to show for them.