Video analytics in the automotive business
Functions of video analytics applied in transport
Our cars are getting smarter every year. Of course, the war between the Decepticons and the Autobots is still far away, but the beginnings of artificial intelligence have already settled in almost every vehicle.
We always seek to provide a greater level of comfort to our life and the economic efficiency of a business. To achieve this, we are rapidly developing new technologies and constantly introducing them to our daily routine. Video analytics is one of the fastest-growing technologies, which uses video data to automatically analyze situations without direct human intervention.
Transport is one of the most prospective areas for video analytics. The implementation of this technology will affect both the auto business and the overall culture and safety of driving.
In the Ivideon company we currently receive many requests from the business for such basic things as passenger counting in public transport, monitoring of driver’s working hours, detecting an unauthorized person driving a car, or the engine starts using Face ID. There still might be certain difficulties in its implementation, which we will discuss later.
But any obstacles are just another impetus to new solutions and better results. In this article, I’ll describe the possible ways of application of video analytics on transport, common difficulties one can face, and decisions to break through barriers.
Functions of video analytics applied in transport
The demand for intelligent video surveillance systems in transport is constantly growing. Such systems help to increase business efficiency, improve public safety, and prevent law infringement.
1. Auto business
ANPR (Automatic Number Plate Recognition)
The first video analytics function that can be used in the auto business is the Automatic Number Plate Recognition (ANPR) system. ANPR is one of the most widespread and well-known video analytics functions designed not only to monitor traffic violations but also to improve the efficiency of customer service.
For example, ANPR solves the issue of the management of parking lots. The system tracks the number of empty parking spaces and informs the driver or security at the checkpoint, therefore, reducing congestions and increasing the efficiency of the parking lot.
Based on the collected data, the system can provide a detailed report on the use of the parking space to make the right business decisions.
ANPR can also be integrated with the payment systems, which will work in the following way. ANPR fixes the parking time, and the money for parking is debited automatically from a linked bank card or other means of payment. This improves the convenience of parking for drivers and reduces maintenance costs for business owners.
In September this year, Moscow launched traffic cameras that capture cars driving through parking spaces. If the driver violates road marking at the parking lot, then the system registers the plate number of his car and issues a fine.
Thus, ANPR makes parking lots not only more convenient but also safer. With such a system, the chance of accidents in the parking areas is significantly reduced.
ANPR improves the service level: the system recognizes the registered plate number and automatically opens the gate. The speed of service increases, the queues disappear, and the cost of human resources decreases (a security guard, who is constantly monitoring the traffic, becomes unnecessary anymore).
One shouldn’t limit the scope of ANPR to just the above examples. The system can also increase the efficiency and profit of car washes, hotels, gas stations, garages, car workshops, auto cafes, multifunctional zones, or enterprises’ access control. If the ANPR system helps to reduce costs and, at the same time, improves the usability of the service, then it helps to earn money.
Another interesting solution for auto business is the Face ID feature. The Yandex company, for example, is actively testing the authorization of the drivers of the Yandex.Taxi service using Face ID technology. The face and voice features of the driver will serve as proof of identity. The greatest advantage of such a key is that it’s always with you, while it cannot be faked, forgotten at home, or lost.
The engine will start only after the identity verification process. The system will prevent the stranger from driving a taxi, which will increase the safety and prestige of the service. This, in turn, will increase the company’s profits and eliminate unpleasant incidents the passengers may face. Within the next three years, Yandex.Taxi plans to invest $52.1 million in taxi security technologies.
DMS (Driver Monitoring Systems)
Driver Monitoring Systems (DMS) can offer another beneficial business solution. Such modules allow the business owner to monitor the undesirable behavior or state of the driver, such as sleep, fatigue, eating, smoking while driving, talking on the phone, distraction from road control, and even the driver’s absence from the workplace.
The system will provide a detailed report on the driver’s working time and will contribute to maintaining the driver’s attention on the road.
2. Driving safety and comfort
ADAS (Advanced Driving Assistance Systems)
As of today, ADAS functions (Advanced Driving Assistance Systems) are widely used in the modern industry of cars. These modules assist the driver in keeping eyes on the road and avoiding accidents.
Already now, they can help the driver to stay in the traffic lane, slow down while approaching the other vehicle, or notify the driver that the movement ahead has begun. In fact, the same analytical modules in their more advanced form are implemented in self-driving cars.
A good example of the ADAS system is the automatic parking function. At the end of August 2020, Ford and Bosch announced that they signed a cooperation agreement within the framework of which they have developed a technology for the automatic parking of cars. The system takes driving control, and the car can freely maneuver and park without the driver’s intervention.
ITS (Intelligent Transportation Systems)
Another key function of video analytics is tracking incidents on the road. Cameras record the appearance of a pedestrian on the roadway or obstacles blocking the path and also detect smoke in the tunnel, congestion, and collisions of cars.
According to the Technical University of Munich (TUM), the latest software developed by one of its departments is capable of reassessing changing traffic situations millisecond by millisecond.
The institution says the software has been conceived to handle very complex traffic interactions, such as those at intersections where pedestrians, other vehicles (which can act erratically), and weather conditions all come into play.
Another example of the use of an intelligent transportation system (ITS) is toll sections of the highways. The system in such areas helps to control and manage traffic, notifies drivers about weather conditions, or any other changes in the situation on the road. Some highways have information boards that, in real-time, reflect various adverse road conditions.
Apple also continues to invest in driving safety and comfort. The company is currently popularizing the Face ID function for keyless access to the car. iPhone owners will be able to create their own virtual car keys, adding them to the memory of their smartphones, and then use them to unlock doors or start the engine.
This access method is more reliable than a key, as, firstly, it minimizes the risk of loss and, secondly, it allows using Face ID to verify the owner’s identity to avoid strangers’ access to the car.
The US State Department, in turn, already uses one of the largest face recognition systems in the world with a database of 117 million Americans, with photos taken from a driver’s license.
This base opens up an opportunity for the development of Face ID technology to verify a person’s right to drive a car using their biometric data. Such technology will help not only to detect a stranger driving a car but also to solve the problem of driving a vehicle by a person who does not have the driver’s documents. Which again improves road safety.
3. Public transport
How can video analytics affect public transport? Intelligent video surveillance systems provide various benefits for passengers, carriers, and transport agencies. We will take a look at a few of the solutions that video analytics offers to improve the global public transport system.
Passenger traffic control. The transport can be overcrowded during rush hours, which may negatively affect the productivity of the transport system and the comfort and safety of passengers.
The video analytics system counts the number of people, sets a threshold, and sends a signal to the driver that there is the maximum number of passengers in the cabin. Moreover, this system will help transport companies to more efficiently arrange transport on routes, for example, to put more units during peak hours or, vice versa, to stop the empty vehicles.
Behavior analysis. Video analytics suppresses unwanted, deviant social and criminal behavior of passengers. Which, in turn, increases the public transport prestige and security.
Security. The camera installed in the driver’s cabin will not only track their actions but monitor the general situation in the cabin. Moreover, the video archive taken from these cameras is the main source of information in case of traffic accidents. In the case of a disputable situation, the video will help to establish the identity of the offender and also track the unaccounted provided services.
Blind zone detection. Cameras can monitor areas that are out of the driver’s field of vision. It increases safety and helps to avoid accidents, therefore, reducing financial losses for companies.
There are also video analytics modules that assist in countering a terrorist threat. They are warning about intrusion into the protected area or violation of the perimeter, as well as detecting fire, smoke, and abandoned objects.
Moscow Department of Transport shows a good example of the video analytics application in the field of public transport. They have recently implemented the FacePay system at some underground stations in the capital. In the near future, the system will actively develop in terms of service for passengers.
The above examples show us how quickly video analytics enters our daily life. Intelligent systems aim to increase the quality of passenger service in transport, which increases the profitability of the entire transport infrastructure.
Difficulties in implementing video analytics in transport
Despite all the advantages of the use of video analytics in transport, there are still some difficulties in implementing the technology.
Quality of internet connection. The internet connection is not always available on moving objects. Therefore, the installed analytical module won’t always be able to give real-time prompts to the driver, or there may be some disruptions in the response speed of the device due to poor connection. Of course, an increase in connection coverage could solve the above problem. As of today, we can offer the use of edge devices with on-board analytics and from time to time synchronization of statistics with cloud storage.
Cost. Video analytics devices should be fairly cheap to make software modules more popular and widespread. Only with an affordable price, it’s possible to place modules of video analytics in every car. Today, it seems to be achievable with the appearance of cheap miniature motherboards like Nvidia Jetson Nano with fairly powerful GPUs.
Infrastructure. Insufficiently developed infrastructure in many countries (primarily low lighting) limits the efficiency of video analytics systems used in transport. Due to poor conditions, the number of false alarms of the system increases, which leads to the extra costs to find out whether this or that incident has actually occurred.
Growth prospects of the global video analytics market
The analysts of Market Research Future predict the growth of the global market for video surveillance in transport. In 2019, investments in this structure amounted to $15 billion, and by 2023 it’s expected that the investment volume in video analytics in transport will almost double. And the compound annual growth rate will be approximately at the level of 14%.
According to MarketsandMarkets primary factors driving the adoption of video analytics solutions are as follows:
1. The increased focus by state organizations to improve public safety.
2. AI systems cover new areas such as enforcing physical distancing, efficient contact tracking, predicting or identifying crowd gathering, identifying offenders, and work time control.
3. Increasing smart city initiatives by states across the globe and the need to reduce crime rates.
4. Cloud technologies, the implementation of which offers numerous advantages over server-based analytics, for example, reducing the cost of infrastructure maintenance.
5. Growth in demand from high-developing economies such as China or India, which are expected to offer huge opportunities for growth in the coming years.
As of today, the video analytics market remains the developing one, diverse and far from being oversaturated. Many companies and studies predict market growth in the world in general and in the auto business sector in particular.
The auto business area, like no other field, can expect great changes and innovations. Technologies will help auto fleet owners or retailers to increase their profit, as they will better understand and control what is happening to their cars, drivers, and passengers.
Urban traffic will become safer, the prestige of public transport will rise, while the number of accidents will fall, and new opportunities will open up to improve the level of customer service.
Overcoming some difficulties in the implementation and spread of technology, video analytics is already becoming more accessible and can solve problems not only in transport but also in many other areas of our life.
I am a tech product manager experienced in innovative products. Launched one of the world's first B2B Cloud Face Recognition platforms which is used now by people all around the globe. Expert in AI/ML, IoT and Web products. Constantly learning. Love to work with great engineering teams