A driver searching for an open space in a parking facility loses more than their own time. Known in transportation research as "cruising for parking," this behavior is the hidden component of congestion in city centers. Measurements vary widely by area and time of day: a field study in China found that between 9% and 56% of traffic was searching for parking, with an average search time of about 6 minutes. Another measurement in central Stuttgart, based on 876 observations, put the share at 15%.

That wide range is the starting point of this guide: a smart parking system is not a product; it is a chain of measurement and intervention. You cannot decide which technology you need without knowing how big the problem is at your facility. In this article you will find:

  • The five layers of the system and which problem each layer solves
  • A clear comparison of ultrasonic, magnetic, and camera-based detection
  • Infrastructure costs that often do not appear in proposals
  • Turning occupancy data into pricing: the 85% rule and the SFpark experience
  • How to build KVKK (Turkish Personal Data Protection Law) compliance into the system design
Cameras, occupancy sensors, guidance displays, a payment station, and a control dashboard working together in a smart parking ecosystem
A smart parking system is not a single device; it is an ecosystem in which the detection, recognition, guidance, payment, and management layers work together.

What is a smart parking system?

A smart parking system is an integrated technology layer that detects the occupancy of parking spaces in real time, guides vehicles to open spaces, manages entry and exit through automatic recognition, and consolidates all of this data in central software.

Academic research states the rationale for smart parking systems plainly: in city centers where parking demand is high, there is no land left to build new parking. What remains is the efficient use of existing capacity and demand management. Smart parking systems are positioned as the most current tool for both goals, and their effectiveness depends directly on how well the vehicle recognition and detector technologies perform.

In practice, this means the value of the system is measured not by how many devices you install, but by the quality of the data it produces and your ability to make decisions with that data. A poorly calibrated sensor network is worse than no sensors at all, because it sends drivers to spaces that are not there and permanently erodes trust in the system.

The five layers of the system

The most common mistake when requesting proposals is mixing up the layers. Under the label "smart parking system," vendors sometimes sell only barrier automation, and sometimes only occupancy indicators. Looking at the five layers separately makes proposals comparable.

1. Detection

Space-by-space occupancy detection using ultrasonic, magnetic, or camera-based sensors. Without this layer, the word "smart" does not apply; a system that only counts entries and exits cannot tell which zone inside is free.

2. Recognition and access

Identifying the vehicle and controlling the barrier: license plate recognition (ANPR), HGS/RFID tags, card access, or a mobile app. Subscribers and visitors are separated at this layer. For details, see our license plate recognition systems page.

3. Guidance

LED indicators, level and zone displays, and entrance totems. This layer delivers the data produced by the detection layer to drivers in real time. It is the layer that actually reduces circulation and emissions inside the facility.

4. Payment and collection

Pay stations, contactless cards, mobile payment, QR codes, and HGS integration. By reducing cash handling, it cuts both queues and revenue leakage.

5. Management software and analytics

Subscription management, rate configuration, shift tracking, revenue reconciliation, and occupancy and turnover reports. This is where the data from the other four layers is turned into decisions, and where most of the return on the investment is generated. Park Expert builds this layer with its parking management software.

These layers connect to one another over protocols such as RS485, CANBus, or TCP/IP. The critical question when evaluating a proposal is this: do the layers come from the same manufacturer, or is it an integration of different brands? In the second case, the contract must state clearly who is responsible for the integration.

Sensor technologies compared

The detection layer is where the budget is debated most, and the choice depends largely on the geometry of the facility.

Technology Coverage Installation Additional data Best suited for
Ultrasonic One sensor per space; cost grows linearly with capacity Ceiling-mounted; practical in covered facilities None; occupied/vacant only Covered and multi-story garages, mid-sized facilities
Magnetic / in-ground One sensor per space Requires cutting into the surface; costlier than ultrasonic None Open lots and on-street applications with no ceiling
Camera-based ~3 spaces per device in single-sided installations, ~6 in double-sided installations Mounted in the middle of the aisle; cabling to fewer points License plate, vehicle type, color; dual use as a security camera High-capacity facilities with a regular aisle layout

Coverage figures reflect manufacturers' typical installation scenarios; aisle width, ceiling height, and lighting conditions change the actual value.

Two points are often overlooked with camera-based solutions. First, the software license is a cost item separate from the hardware, and some solutions use a one-time license while others use a subscription model; the comparison should be based on three-year total cost of ownership. Second, accuracy at night and in bad weather depends on the camera itself: loading license plate reading software onto a camera without built-in infrared illumination and a global shutter sensor produces an installation that is useless in the field.

Budget for the calibration period

Once hardware installation is complete, you need to test that every sensor works correctly, that the displays are synchronized, and that the data flow is uninterrupted. This stage typically requires a 1-2 week calibration period. Projects that schedule go-live for the day installation finishes end up opening with faulty data.

Vehicle recognition: license plate, HGS, and RFID

The recognition layer in Turkey has become noticeably simpler in recent years. OGS, which relied on an in-vehicle device, was fully discontinued in 2022 and replaced by HGS, Turkey's electronic toll collection system, which uses an RFID sticker on the windshield. Today, parking facilities use three methods side by side.

License plate recognition (ANPR)

The camera reads the plate, matches it against the database, and opens the barrier. No consumables such as cards, tags, or remotes are required.

  • No consumable costs
  • Every pass recorded with a photo and timestamp
  • Accuracy is highly sensitive to camera quality

HGS tag

A long-range UHF antenna reads the tag, and the vehicle passes without stopping. The fee is charged to the HGS balance. For details, see our HGS parking system page.

  • Strong in facilities with many subscribers and fleet vehicles
  • Eliminates cash handling entirely
  • Requires antennas and tags

Card and mobile access

Barrier control by RFID card or mobile app. The simplest option to install and the lowest-cost option to manage.

  • Common at residential complexes, office towers, and schools
  • Lost cards create an ongoing cost
  • Identifies the user, not the vehicle

HGS adoption on the parking side continues to grow. In Bursa, BURULAŞ rolled out the HGS Park payment system as an additional option alongside existing methods at 29 parking facilities operated by BURPARK: 12 on-street and 17 off-street lots.

The right combination depends on the facility's traffic profile. In a subscriber-heavy office tower, HGS or cards provide frictionless access; in a shopping mall dominated by same-day visitor traffic, distributing tags is impractical, so the return is limited and license plate recognition comes to the fore. Most serious projects install both.

Cross-section of a multi-story garage showing sensors detecting occupied and vacant spaces and a driver being guided to an available space
When level-by-level occupancy data is fed to digital displays, drivers are guided straight to an available area without having to search.

Guidance and occupancy indicators

The guidance layer is where detection data reaches the driver, and it is where you make the most visible contribution to the customer experience. LED indicators mounted above each space show its status by color: green for vacant, red for occupied, navy for accessible spaces. With RGB LEDs, different colors can be assigned to VIP or reserved spaces.

Displays at level and zone entrances show how many spaces are free on each level, so drivers do not have to enter a full level. This directly reduces unnecessary circulation inside the facility and the congestion that comes with it.

The guidance layer has an often-overlooked side benefit: accessible parking spaces become monitorable. Unauthorized use can be seen and reported in real time through the system. For both regulatory compliance and corporate reputation, this is a level that manual enforcement cannot reach.

From data to pricing: the 85% rule

This topic is almost entirely absent from Turkish-language sources, yet it has the greatest impact on the return of a smart parking investment. Sensors give you occupancy data. So what do you do with it?

The widely accepted approach in transportation economics is to eliminate cruising for parking by pricing so that occupancy stays below 85%. The logic is simple: as long as a few spaces remain free on every street or in every zone, drivers do not have to circle. When occupancy hits 100%, search time and circulation rise sharply.

San Francisco's SFpark project applied this approach in the field, periodically adjusting prices by area to bring occupancy into a target band of roughly 60-80%. A modeling study calibrated with SFpark sensor data found that each price adjustment moved a block segment 0.1-0.2% closer to the target band on average and reduced cruising by more than 5%. Cities such as Los Angeles, Calgary, and Seattle have also moved to similar demand-responsive pricing.

What it means in practice for private facilities

You may not be a municipality, but the same logic applies. A shopping mall parking facility whose occupancy runs above 90% at peak hours is not increasing its revenue: every vehicle it turns away is lost sales, and those who do get in lose time circling. With sensor data in hand, tiering rates by time of day shifts long-stay, low-yield parking to off-peak hours.

This is the real source of a smart parking system's return: not the hardware itself, but the pricing decisions made with the data the hardware produces.

Hidden cost items

The price of a parking system has four components: hardware, software, installation and infrastructure, and support. The biggest cost driver, however, is not the brand but the number of entry and exit points and the level of automation. These are the items most often missing from proposals:

1. Concrete cutting, trenching, and cable pulling

In a new facility, this work is planned during construction and costs little. In an existing facility, the same work means additional labor and temporarily closing part of the facility.

2. Electrical infrastructure and protection

Power lines, grounding, and surge protection. In a field installation without lightning protection, a single event can take out every camera and controller board.

3. Software license and renewal

A one-time license, or an annual subscription per vehicle or per user? Any comparison made without calculating the three-year total is misleading.

4. Maintenance and repair response time

If the contract does not define a response time (SLA), one Saturday afternoon with a faulty barrier locking the exit will cost the entire investment its reputation.

5. Staff training and handover

Training facility management and security staff in reporting, maintenance, and fault tracking is part of the installation. Untrained staff will run even the most advanced system in manual mode.

The good news is that retrofit projects do not require starting from scratch. Infrastructure items such as barriers, posts, and cabling can often be kept, with only the reading, payment, and software layers upgraded. Documenting reusable components during the site survey lowers the budget significantly.

KVKK compliance

License plate data and vehicle images may qualify as personal data. That makes every operator that installs a smart parking system a data controller, and compliance must be part of the design, not a document added after the system goes live.

  • Privacy notice. Processing of data through cameras and license plate recognition must be clearly announced at the parking entrance and in the facility's information channels.
  • Legal basis for processing. In practice, processing carried out to ensure security and physical premises security is based on the legal obligation and legitimate interest grounds in Article 5 of KVKK. The specific basis must be identified.
  • Retention period. How long records are kept must be defined in advance, and automatic deletion must run at the end of that period. When an incident occurs, only the relevant segment is set aside and kept for the duration of the legal process.
  • Camera angles. Cameras should be positioned to cover only the area the purpose requires; residential entrances, balconies, and windows should be kept out of the field of view.
  • Access authorization. Who may access the recordings, and on what grounds, must be defined, and all access must be logged.

There is a tension here, and it should be named honestly: as we discuss in our parking operations guide, the operator is legally liable if a vehicle is damaged or stolen and must keep records to be able to defend itself. From a KVKK perspective, the same record is data processing that must be limited. The right balance is struck through purpose-bound, time-limited retention; keeping no records at all is as wrong as keeping them indefinitely.

Selection table by facility type

The same system does not fit every facility. The table below is a starting framework for setting priorities by facility profile; the final decision is made after a site survey.

Facility Dominant traffic Priority layer Recognition method
Shopping mall Same-day visitors, heavy peak hours Detection + guidance (experience and capacity) Mainly license plate recognition, supported by HGS
Hospital Mixed; long stays for patient companions, short stays for outpatient visits Guidance + tiered rates License plate recognition; tags/cards for staff
Office tower / business center Subscriber-heavy, morning and evening peaks Recognition + software (subscription management) HGS or card; priority on frictionless access
Residence / residential complex Residents' vehicles + visitors Recognition + security integration License plate recognition; temporary permits for visitors
On-street / open lot High turnover, short stays Detection (magnetic) + payment License plate recognition or mobile payment

Conclusion

A smart parking system is a decision chain, not a list of devices. The detection layer produces the data, the recognition layer manages access, the guidance layer carries the data to the driver, the payment layer closes the leaks, and the software layer turns all of it into decisions. A project that skips a layer does not get its money's worth.

Sensor choice depends on the facility's geometry: ultrasonic stands out in covered and mid-sized facilities, magnetic in open areas, and camera-based solutions in high-capacity facilities with regular aisles. Compare proposals on three-year total cost of ownership and make sure concrete cutting, electrical infrastructure, license renewal, the SLA, and staff training are included.

Most importantly: the return on the investment comes not from the hardware but from the pricing and capacity decisions you make with the data it produces. A facility that starts measuring occupancy and redesigns its rates around that data within the first year usually gains more than the system itself cost.

Frequently asked questions

What is a smart parking system?

A smart parking system is an integrated technology layer that detects the occupancy of parking spaces in real time with sensors, guides vehicles to open spaces, manages entry and exit through automatic recognition, and consolidates all of this data in central software. Its purpose is not just convenience but the efficient use of existing capacity and demand management.

Should I choose ultrasonic sensors or a camera-based system?

With ultrasonic sensors, every parking space gets its own sensor; cost scales linearly with capacity, but detection is simple. In a camera-based system, a single camera can monitor about 3 spaces on one side of an aisle, or about 6 in a double-sided installation, and it generates additional data such as license plates and vehicle type. Cameras are usually the more rational choice in high-capacity facilities with a regular aisle layout, while ultrasonic sensors tend to make more sense in irregular, lower-capacity areas.

How long does it take to install a smart parking system?

It depends on the existing infrastructure and the number of entry and exit lanes. A single-lane facility with infrastructure already in place can be commissioned quickly, while projects that require concrete cutting and trenching take longer. After the hardware is installed, a calibration period of typically 1-2 weeks is needed to synchronize the sensors and displays.

Is license plate recognition in a parking facility subject to KVKK?

Yes. License plate data and vehicle images may qualify as personal data, and the party operating the system becomes the data controller. The privacy notice, a clear legal basis for processing, a defined retention period, and data security measures should all be planned together with the system installation.

Is OGS still used in parking facilities?

No. OGS, which relied on an in-vehicle device, was discontinued in 2022 and replaced by HGS, which uses an RFID sticker attached to the windshield. On the parking side, HGS tags and license plate recognition are now used together.

Related content

This content is for general information purposes only. Sensor coverage figures, installation times, and cost items vary by manufacturer and site conditions; final planning should be based on a site survey. Data on cruising for parking and occupancy targets is based on field studies in the transportation economics literature (Shoup 2006; Hampshire & Shoup 2018; Millard-Ball et al. 2014; Zhu et al. 2020). For KVKK compliance, obtain a legal assessment based on your own data inventory.