The problem these systems address
Drivers looking for a space add to congestion, and cities have limited curb space to share between residents, deliveries and visitors. Parking and traffic sensors aim to give two kinds of information: whether a particular space or lot has room, and how many vehicles pass a point and how fast they move.
Detecting an occupied space
Several technologies are used, and the choice depends on whether spaces are on a street, in a lot or in a garage.
- In-ground sensors are small units set into the pavement of a single space. Many use magnetic sensing, because a car’s metal changes the local magnetic field, and some combine it with other signals to reduce false readings. They report occupied or free by radio, usually over low-power networks. Their battery life is a central design issue, since digging up pavement to replace a battery is expensive.
- Overhead ultrasonic or infrared sensors are common in garages, mounted above each bay and often paired with a light that shows green or red.
- Cameras can watch many spaces at once. Software identifies whether each marked space holds a vehicle. One camera can replace many in-ground units, but it needs a clear view, adequate light and a decision about what footage is kept.
- Entry and exit counters track vehicles in and out of a lot. Subtracting one from the other gives the number inside, though counting errors build up over time unless the figure is periodically reset.
Each approach has trade-offs. In-ground units are not affected by weather or lighting but cost labor to install and can be harmed by road works or snowplows. Cameras are flexible but sensitive to conditions. Counters are cheap yet can drift.
Sensing traffic flow
Traffic flow measurement answers different questions: how many vehicles, how fast and what types. Typical tools include:
- Inductive loops cut into the road surface, which detect the metal of a passing vehicle. They have been used at signals for a long time.
- Radar units mounted beside or above the road, which measure speed and count vehicles in several lanes.
- Video analysis, where software counts and classifies vehicles, such as cars, trucks, bicycles and pedestrians.
- Pneumatic tubes laid across the road for temporary counts.
- Anonymized travel-time data from connected vehicles and phones, which can show congestion over a wider area without any hardware in the road.
Signals can use this data, adjusting green times to the queue instead of following a fixed timer. Cities also use historical counts for road safety reviews and for deciding where to add crossings or lanes.
Getting the data to drivers and planners
Space status goes to a central system, which updates signs at the entrances to car parks, mobile apps and enforcement tools. A driver sees a count of free spaces on a sign rather than a tour of full lots. A city can also adjust prices by area and time, charging more where spaces are scarce, though whether that suits a given place is a policy decision, not a technical one.
Planners study occupancy by hour and day of week to learn when spaces are really full, how long vehicles stay and whether time limits are working. Over time, that record can inform urban planning choices, such as whether loading zones are well sited. The patterns are stored as time series data and can be explored in sensor dashboards.
Limits and honest expectations
Accuracy is not perfect. A motorcycle might be missed, a large truck might trigger two spaces, and a magnetic sensor can be confused by nearby heavy vehicles. Operators check readings against occasional manual surveys, and data quality problems should be expected.
Coverage matters. Sensing only some spaces gives a partial picture. Drivers may not trust a sign if it is often wrong, so reliability is more important than coverage in some cases.
Maintenance is ongoing. Sensors fail, batteries run out and gateways lose power. IoT gateways collecting the signals need monitoring just as much as the sensors do.
It does not create space. Better information may reduce circling, but how much depends on the city and the level of demand. Guiding drivers to free spaces is sometimes said to reduce traffic, but the effect varies and is hard to measure.
Privacy and fairness
Plate-reading cameras, used for enforcement or payment, can record where a vehicle was and when. That is personal information, so clear rules on what is recorded, how long it is kept and who can see it are important. Sensors that only report occupied or free carry fewer concerns than any system that stores vehicle identity. Planners also need to consider whether pricing or enforcement is applied fairly across neighborhoods, since poorly designed systems may weigh more heavily on some groups.
A sensible way to start is a small pilot in one area, comparing sensor results with counts taken by hand before expanding.