Automation in production, business and the public sector is impossible without a thorough analysis of the processes taking place in the organization. Digital transformation provides the best tool for these purposes – video analytics. It is a new source of data for business development, helping companies get answers to the following questions:
What can be done to make the business more profitable?
-
how to optimize production?
-
how to save money?
-
how to increase income?
-
how to improve performance?
The key point is the choice of the optimal service that will cover all the needs of the customer and help the organization reach a new level.
Five truths you need to know to select the optimal system
Truth #1: Video analytics is not video surveillance
It is important to understand that video analytics and video surveillance are different solutions.
Video surveillance is a reactive system. It reacts to what is happening and provides information after the fact.
NB! The simplest video analytics, some of the old systems that complement video surveillance systems, are also reactive systems.
A comprehensive advanced video analytics system is a proactive system. It provides not only alarm notifications, but other information, in particular graphs, reports, statistics, which can already be used to build predictive models, and in the system itself (for example, the likelihood of incidents or performance degradation).
Video analytics platform customers can be divided into three categories:
50% do not understand what video analytics is, they confuse video analytics and video surveillance.
-
30% do not fully understand what video analytics is, perceiving it as a tool for notifying about any actions (for example, about entering a hazardous area in a production facility) or for recognizing employees’ faces when they enter an organization.
-
And only 20% understand that video analytics is a tool for implementing digital transformation processes, data analysis and business analysis within the company.
Such a difference in perception arose from the fact that video surveillance has existed on the Russian market for a very long time – almost 100 years: the first video surveillance system in the USSR was launched in 1926.
The giants of the domestic industry were the first to calculate the economic benefits of integrating the product into the company’s infrastructure – they train functional customers to make new management decisions based on this information.
For example, a pipe plant, whose finished product accounting system is built using video analytics: it counts produced pipes and classifies them according to certain parameters (such as diameter), separates quality products from defective products, and generates and provides an analysis of production efficiency. A complete statistical report gives managers the information they need and an understanding of when performance falls and when it rises, on what factors these processes depend.
NB! A distinctive feature of the video analytics system is that it has clear and understandable tasks laid down at the design stage. For example, cameras monitor a specific area in a warehouse and analyze inventory and product receipts.
It should be borne in mind that video analytics may include boxed modules that are well-established and somewhat standardized. But for effective work, it is necessary that the system be built into the production processes of the organization. Customization always requires adjustment to the regulations and regulatory documents of the organization in terms of integrating the technical solution into the customer’s business processes.
Truth #2: Data analysis is different from business analysis
NB! It is important to distinguish between systems that automatically recognize people or objects in video according to specified parameters without further analysis, and analytical systems that analyze the data received.
Most video analytics systems are a complex that replaces a conditional operator. Such systems themselves are engaged in video surveillance, notify about incidents and extract the necessary data. For example, analysis of compliance with safety regulations at work: the systems record the absence of work helmets, gloves or seat belts, signal a danger, and provide an upload with data for a report.
To solve larger tasks, you need an additional rate of a business analyst in the structure of the company. This is necessary to convert the data received from the system into a format that is necessary and understandable to the business.
An alternative solution is a “smart” platform that can automate and improve the workflows of complex video analytics systems. Data from CCTV cameras are loaded into the information system, where they are processed and formed into statistics, which are displayed on the dashboard in the form of graphs and diagrams. Thus, the functions of not only the operator, but also the business analyst are automated.
For example, a system related to recognition of wagon numbers for an oil company. The train passes under the camera, the system recognizes and fixes it, automatically adds data to the reporting document. This is a simple data upload, without the necessary analysis. But if the system capabilities are expanded to create a complete analytical report required by the customer, then, in addition to recognizing the train itself, the analysis will take into account such data as the number of cars (with photographs of each of them), route sheets, and direction of movement. Such a report does not need additional processing by a business analyst. The system, created and customized to the needs and scenarios of a particular customer, will do all the routine work automatically, and the customer will receive information ready for use.
Truth #3: There are no clear standards in the market for video analytics
The market for video analytics is just emerging in Russia. Such systems solve fundamentally new tasks, and organizations do not always understand what they can count on when integrating a new product, what opportunities they will receive and what functionality they can request. In 90% of cases, the initial statement of the problem differs from what the client actually wants to receive.
For example, recognizing car models from traffic cameras and keeping statistics is an innovative idea. People didn’t sit on poles and count cars before the introduction of video analytics. And today this solution provides information to a variety of structures in the context of their specific goals, objectives and requests.
Thus, everything depends on the business processes of the customer. Therefore, in order to eventually get a high-quality and useful custom solution, it is necessary to understand what kind of analysis and what kind of data is needed.7
Truth No. 4. There are different classes of systems on the market
Due to the lack of clear standards and technical specifications, different functionalities are invested in a product called Video Analytics, which leads to the appearance on the market of systems of various classes under the same name.
For example, recognition of wagon numbers for an oil company. The customer has been looking for such a solution on the market for a long time and eventually chose the “smart” RTMIP platform, which provided a recognition level of 95% at the piloting stage and 98-99% after the system was finalized to meet the specific needs of the organization: the movement of cars in daylight, in the dark, in difficult climatic conditions, conditions of air pollution and poor visibility in the region.
Customers set tasks for implementation, which are ranked from simple to complex. Similarly, you can distribute the classes of video analytics systems on the market. From solutions that can close a minimum of requests, to custom products focused on the most difficult tasks.
NB! The cost of a solution does not always determine its functionality!
Customers have to choose from dozens of solutions. And the time spent and the integration of products that will not suit you in the future are associated with financial losses.
Truth #5: An open platform eliminates endless migrations
The best option to save time and money is to seek help from a trusted partner, an IT integrator. Solution providers typically employ teams of highly qualified architects and engineers who perform technical audits of vendor solutions and products on the market. But when integrating a video surveillance system, it is necessary to take into account not only the characteristics of the supplier and its product, but also the conditions, requests and expectations of the customer company.
So, if the customer has “ideal” conditions – full illumination, no smoke, dust, etc., a “boxed” solution will be quite enough. And in cases of “difficult” projects, when the customer comes with a complex task that previous suppliers could not solve, it is necessary:
qualified team of architects of computer vision systems;
-
development customization;
-
a powerful video analytics platform that can flexibly customize scenarios for specific customers.
NB! If the platform is open, you can not limit yourself to approved recognition scenarios, but edit algorithms, create new ones, and at the same time combine the functionality of different vendors.
For example, it is possible to assemble solutions for technological video analytics in the field of industrial security and protection on one platform. At the same time, information flows will be divided into responsible departments of the enterprise.
The IT market is very dynamic – new products, modules are constantly being released, the list of services is being updated. The advantage of an open platform is that when a more advanced solution appears, you can simply replace one of the components of the platform, and not demolish it entirely. This approach has many advantages:
saving the customer’s budget;
-
the customer purchases digital archives;
-
the interface of the platform does not change, it is comfortable to work with it;
-
there is no need to retrain staff, as is the case with a complete migration from one solution to another.
Turnkey system integration
The video analytics system is designed for specific customer tasks. This process includes the selection of cameras, zones and equipment installation points.
The turnkey video analytics system integration engineering project consists of seven stages:
Audit of the customer’s current infrastructure. For example, whether there are cameras at the enterprise and whether their use in the system is possible.
-
Designing locations for installing new cameras.
-
Installation supervision and organizational and technical work, supervision by the equipment supplier for their implementation. For example, when installing equipment.
-
Commissioning work performed on the installed equipment. For example, putting an object into operation.
-
Fulfillment of warranty obligations, as well as assistance in training personnel for the operation of the facility.
-
Improvement of the system if necessary.
-
Technical support of the system.
NB! When concluding a contract, it is better to prescribe what tasks and with what accuracy the video surveillance system should solve. This guarantees not just the installation of cameras for show, but the fulfillment of obligations that correspond to the business processes of the customer, on a turnkey basis.
Great advice to the customer
Video analytics only occupies its own niche in the Russian market. The increase in demand for such solutions entails an increase in offers from various vendors. Photos, commercial offer or videos will not show the pitfalls of the product, will not help to understand whether this solution is suitable for the customer.
NB! The only way to evaluate a video analytics system without losing your budget is to test the system live, on cameras installed by the customer.
It is better to design a video analytics system, install cameras, and only after that start choosing software that will allow video analytics.
During live testing, the customer understands:
-
the stage of system readiness, because the contractor can only have a presentation of the solution and promises to implement everything;
-
how the system works under certain conditions;
-
how reliable information is given;
-
whether such a system is suitable for the business processes of the enterprise, or another option is needed.
Over the past two or three years, the level of technical literacy of customers has grown significantly. Attention to product details, implementation nuances and a proactive approach will help them make an informed choice and make the best decision.
