How to Implement New Parking Slot Detection System using Prototype

When imagining a busy day in a crowded city, people will never inspire to go outside because they have to spend an extra amount of time, effort and fuel while finding a suitable and available parking space. Nowadays, many buildings have solutions for free parking space indication. Thus the technology of parking detection systems and management usually sufficient for parking lots inside of the buildings. However, these solutions and technologies are not enough to satisfy the parking requirement when considering exterior parking lots and especially for heavy crowded areas. Our iTech Mint blog proposes solutions to this parking problem. This parking issue directly impacts to increase traffic problems when people have to take a huge effort to find a place to park their vehicles. It causes traffic jams and air pollution in city centers. Hence, we compare and contrast current technologies by investigating issues such as power consumption, communication approaches, and user-friendly dashboards, then introduce the brand new solution for this problem. Let's have a quick look at the components we propose to use for the project.

A system for detecting a parking space
Bluetooth low energy communication
Arduini based sensor devices

The solution is deployable and scalable with the above-mentioned components accomplish by some kind of a supercomputer (Raspberry Pi) collects all the data which is capturing by IP cameras and upload into the cloud-based database system. The data will available for drivers and they can view the information via their mobile phones and they can decide where 'I'm gonna park my buddy'.

Existing technology behind the scene >

Although there are many existing parking solutions available in city buildings, the requirement is not at the satisfied level. Let's see a common solution in brief. A parking area has red/green light indicators. Drivers can identify the parking space is currently used or is free for parking. This kind of solution can be easily implemented by using the dark area inside a building and electricity can be provided by existing infrastructure. However, when considering the exterior area, we cannot use light indicators because drivers do not see these lights during daylight. Therefore, the light sensors should connect to a navigation dashboard or smartphone app for the indication of parking occupancy details.

As we mentioned earlier, we have compared and contrasted those existing technologies and going to make the parking system architecture. Furthermore with the growth of IT, finding an available safe parking space is getting much easier. We can find plenty of cost-effective sensors and indicators with low power consumption. By using these devices, we can create a LAN/WAN network. Then devices can communicate with each other and provide very useful information about free parking space or utilization of the parking capacity. The city planners and government can review these historical data and analyze them to develop their cities.

Briefly looking into methods which can detect a free parking slot >

Now let's see what are the methods we can detect a free parking space. It can be detected visually by using a video camera. The system uses statistical models for predicting whether there is a vehicle in a parking lot. There are few challenges when implementing the solution such as those cameras and image processing devices are more expensive. The system gets slow on cheap microcomputers because of the high power required for the computation of detection. The system wouldn’t work with battery-powered low power devices that communicate wirelessly. Another solution is to use less sophisticated sensors such as ultrasonic sensors which can measure the distance to the objects in parking buildings. Although the ultrasonic sensor works accurately, its sensitiveness goes bad and not function properly under considerable environmental conditions such as snow or rain. Ultimately, this sensor solution also negative for exterior parking. However, the system can be accomplished with other sensors such as infrared which doesn’t influence by bad weather conditions.

Infrared is the best option for exterior parking sensors >

The combination of ultrasonic and passive infrared sensors reveals a great deal for optimized solutions. The communication among the sensors can heal by WIFI connection to a microcomputer Raspberry Pi. The data captured by sensors stored in a cloud. Drivers can review the data and find an available space for parking and pay bills online. If the driver parks in the wrong parking lot, the system would trigger a notification to the driver's smartphone. Infrared can combine with magnetic sensors to provide an accurate solution. In advance, the exterior sensor devices can power up with solar energy. The project can also be separated into two modes as exterior magnetic sensors and interior ultrasonic sensors while the system uses Bluetooth low energy for communication. Further, the solution and the system can use RFID sensors for the identification of vehicles and the Arduino device as a control unit, sending data to the cloud server database.

Let's see some other solution for free parking >

Heavy crowded but technology-rich cities use sensors mounted in vehicles to detect free parking space. When the vehicle is reaching the city area, the car sends a request to the parking system automatically, then the updated database in the cloud uses an app to send a notification to the drive about the information with regards to the parking space and available location. This solution also has some boundaries. Vehicles have to install the sensor device. Otherwise, while the driver is moving to a free parking space, meantime another vehicle without having the sensor can take the spot.

Reservation based systems >

Some cities have reservation bases systems. The drivers need to be registered in the domain. Once they need a free parking space, they can search online and reserve the spot. The reservation system also has some practical difficulties. The system relies on the discipline of drivers to provide accurate data. If drivers are entering the wrong information to the system, the system cannot provide optimal utilization. For instance, if a driver doesn’t know how long he/she is going to use the parking space, then the parking lot reserved for a longer time period that is necessary. However, for commercial parking lots, drivers have to pay for the entire time. The problem in that manner is other drivers have to wait until checkout time. Therefore, the reservation parking solution cannot be used for real-time occupancy information. And also the system is closed for unregistered drivers.

When considering the network connectivity among sensors and other components, the wired connection is problematic in exterior areas. Therefore, we can use wireless communication which is more scalable and suitable. However, WIFI and Bluetooth technologies required high power and it consumes too much power. Hence, these common wireless technologies are not suitable for battery-powered or energy harvesting sensor devices with a limited amount of energy. Bluetooth technology has been developing and the latest enhancement is that there is more power-efficient Bluetooth low energy technology.

When considering a visual-based parking solution, the detection of a free parking space doesn’t suitable for either low power or power limited infrastructure due to the high performance required for visual-based solutions. Therefore, most commercial solutions are based on simpler sensors such as the combination of infrared or ultrasonic. However, the commercial parking solutions are not exchanging their information with different parking systems. Drivers have to install several applications in their smartphone because the database doesn't synchronize with different commercial systems. Let's assume a driver is going to the town and looking for a parking space. He/she has to open up several applications and search for a free slot. This kind of solution totally relies on the reservation database to provide information about free parking space. If we have a centralized database, the results will accurate. Hence, we come up with a solution that makes it easier to find a free parking slot. The sensors we propose is not too much expensive, widely available and well-known devices which use open source software that doesn’t charge license fees. The solution implies high scalability due to everyone can use a low-cost system and application to detect a free parking space. This solution can be implemented for personal, commercial and street-based public parking. The system would merge with existing parking systems as well. In the next chapter, let's have a quick look at the system and experimental procedure for further discussion.

According to the analysis we have tested for existing parking systems, we summarized the requirement for the proposed system.

Parking space occupancy detection - The system can monitor and track the occupancy of individual parking spots.
Visualization of free parking space - If the system has a visualization where the free parking slot available, then the system can send a notification to the driver where the vehicle can be parked.
Parking statistics - The system can analyze historical data.
Wireless communication – To reduce the cabling cost. Wireless connection between sensors also can enhance scalability and reduce infrastructure cost to implement the solution.
Low energy consumption – Selected sensors can run n battery power for years.
Easily available components – The sensors use for the project are widely available. For the application perspective, the system uses open source software without license fees.
To implement an error less solution, we guide you to select the best appropriate combination between the sensors such as infrared sensors to use for exterior areas while ultrasonic sensors use for interior areas.
Then you have to create a user-friendly android application that presents detected information for drivers. This information should be sent to upper management such as if the system implemented at a local parking-lot, the statistical details would send to administrators. The global parking statistics have to send for municipalities.
The connectivity solution is provided by Bluetooth low energy between peripheral and central devices. When selecting the sensors, we strongly recommended using well-known common sensors because the guidelines, software, and extensions are freely available. After gathering those requirements, let's see the system architecture and the components which illustrate in the below diagram.

The system architecture and its components >

In the above illustration, we can see there are two kinds of devices. The devices which use to collect data from sensors and their responsibility is to send data to central devices. This type of device are called as peripheral devices and we used Arduino UNO in our project. When considering a central device that we met another device type in the project requires higher computer power and memory space to store all the data sent by peripheral devices. We used Raspberry Pi 3 due to its high performance as the central device.

Arduino is a popular microcontroller that comes with different versions. It can collect data from various sensors and has many extension shields that can connect with other devices. We propose Arduino because it is cheap and easy to use. The power consumption is low and it can communicate with the central device represented by Raspberry Pi 3 via wireless technology. Raspberry Pi is a microcomputer which has many useful features. It can connect to the internet via Ethernet or WIFI. Actual Raspberry Pi is some kind of simple computer which operates with a Linux based operating system. Therefore, SQLite3 DBMS can be installed. The central device sends information to the cloud and the data is stored in the backend cloud database server. Then drivers are able to access the information in the cloud and smartphone application helps to find the free parking space by analyzing the information. In the next chapter let's discuss how the data is collecting and storing in a cloud database server. If you have any doubts, please don't hesitate to comment below.

Cheers!