How are small and medium players finding new market opportunities in second half of smart security?

Image source @panoramic vision

Text 丨 Polar body of brain

The water is big and fish is big, perhaps this sentence doesn't quite fit description of today's smart security industry.

In connection with constant improvement of city management system, safety issues in cities are becoming increasingly important. From original public security, traffic safety and key industry security such as finance, campuses and hospitals, it has gradually developed into areas such as parks, communities and families, as well as personal security areas, and a complete scene security system is being formed.

The development and application of AI technology has opened up new opportunities for modernizing security industry. Similarly, security industry is also a mature area where AI technologies, especially machine vision and knowledge graph technologies, develop first and develop on a large scale.

According to China Security Industry Research Report 2019, total output value of national security industry in 2019 was 826 billion yuan, with an annual growth rate of 15%. Intellectual security plays an increasingly important role in it. In 2020, market value of intelligent security market is estimated to exceed $100 billion, of which AI+ security software and hardware market will reach $45.3 billion.

Industrial development of intelligent security can be roughly divided into six stages: "algorithm, computing power, data, solution, implementation, operation and maintenance."

The first three stages can be called first half of "technical support". Whether it's flagship security vendors, integrators or Internet companies, AI companies, and cloud service providers, they're all ready to complete technical layout. The stage can be called second half of "scenario application", which involves introduction of products and technologies into application of specific landing scenarios and requires tougher competition and more realistic tasks. The curtain is about to open.

The characteristic of security industry itself is that market is huge, but scenarios are highly fragmented, so it can not only develop security giants such as Hikvision and Dahua with tens of billions in revenue, but also support tens of thousands of small and medium-sized security enterprises .

Among players who have embarked on path of intelligent security in recent years, we can not only see traditional security experts such as Hai Dayu and Internet giants such as BAT, but also see latest high-profile launch of new ICT giant Huawei .Smart security brand - Huawei HoloSens (good hope), new AI stars such as SenseTime, Yitu, Megvii, and hundreds of companies ranging from basicand ending with terminal solutions.

It is possible that competition in field of smart security will be very fierce. We will definitely see a desperate fight for projects To G and To B with hundreds of millions of allowances, and we will definitely see a price war that continues traditional competition in security market, and we will also often see top players fighting openly and covertly. ..

However, after these usual déjà vu narratives, we hope to delve deeper into intellectual security industry. What new market opportunities can be opened up through technological upgrades and scene immersion? In this gluttonous feast in second half of smart security, in addition to qualifying competition of big players competing for guest of honor, will those small and medium players who are eager to try can feast on it?

Cross AI's "three mountains" and find point of technical strength

With development of technologies such as artificial intelligence, big data and cloud computing, security is moving from traditional video surveillance to intelligent security. Among them, artificial intelligence technology, especially machine vision technology, is widely used in field of video surveillance to achieve transition from "clear vision" to "understanding" and "decision-making ability".

In security realm, AI technology itself is breaking through "three mountains" of computing power, algorithms and big data, helping security industry make transition from "weak AI" to real strong security intelligence.

Increasing computing power depends on improving computing capabilities of AI chips. However, low processing power chips built into conventional cameras make perception and recognition performance unsatisfactory.

Algorithm updates must be supported by a separation of hardware and software and an open ecosystem of algorithms. The traditional camera replacement algorithm requires cameras to be replaced at same time, so after changing algorithm, this is equivalent to rebuilding entire network. If AI ​​camera supports software-hardware decoupling, and algorithm is loaded on demand, it can quickly adapt to business changes and significantly reduce construction re-occurring costs. Advances in deep learning algorithms and an open algorithm learning platform can provide rich algorithms for end-to-end multi-scene recognition.

Data integration requires standardization and interconnection of multidimensional data. The standardization of multidimensional data comes from structured analysis of video data, and multidimensional data is adapted to improve AI recognition rate. Data interoperability requires coordination of device, edge, and cloud data.

Last year, 2019, “three mountains” collapsed one after another,facilitating renewal of intellectual security industry.

In terms of computing power, processing power of main AI cameras completes transition from low computing power of 0.66T to high computing power of 4T, 8T, or even 16T. This will greatly improve foreground and background intelligent processing capabilities, as well as greatly improve performance of face, vehicle detection and intelligent analysis capabilities.

At level of algorithms, various algorithms are developed to deal with complex security scenarios, including multifunctional extraction and recognition, tracking between cameras, intelligent analysis of device-cloud collaboration, etc. Developing and iterating algorithms in separated scenarios provides a space to showcase to many small and medium sized AI companies.

At data level, cloud computing and collaboration between cloud and edge devices have become major trends. Relying on cloud platform and big data, in addition to terminal software and hardware separation, hardware, platform, data and algorithm applications will be hierarchically separated, so that a set of intelligent security systems can be provided by several professional suppliers to provide technical support for module, providing space for development of industrial diversification.

The official commercialization of 5G technology in 2019 and its popularization in next few years will also be a strong incentive for development of intellectual security. The wide bandwidth of 5G is consistent with high-definition video transmission, low latency capability is consistent with real-time monitoring control response, and mass connection provides multi-dimensional perception capability for intelligent security.

The combination of 5G and AI will greatly enhance effectiveness of intelligent security in perception, analysis and decision making, and at same time greatly enrich application scenarios in security industry, transforming intelligent security from existing public security, transportation and other specialized industries. The industry is expanding into richer B-end and C-end industry scenarios.

In this process, in addition to direct benefits of operators, cloud service providers, and chip and equipment manufacturers with their own natural advantages, new start-ups, integrators, and AI channel providers can also benefit from technology, ecology, and creating your own exclusive benefits on channel and enter second half of scene application track.

The scope of application is sinking, and ecological niche of industry has been determined

In terms of industry and business scenarios related to smart securityIndeed, non-civilian areas such as public transport systems and public safety systems are by far most important and largest market areas. For example, in 2018 public safety and transportation projects accounted for almost 40% of total market share.

These areas of AI security are characterized by large scale, significant profit margins, and high technical requirements. Accessible "high barrier areas".

The second largest market area is B-end security market, which includes divisions such as smart finance, smart public (park) areas, and smart campuses. In almost all areas of activity, many traditional security companies and AI manufacturers have been actively involved in game, and even developers like Vanke are already creating their own security platforms to create intelligent communities with their own advantages. traditional public security field, it also engages in access control, property management, building intercom, license plate recognition and other activities through active acquisitions.

The third largest segment is C-end market. Due to overall technical complexity, this has become a segmented area that AI security startups and any hardware manufacturers are trying to participate in. In early stage of competition, when good and bad people were mixed, participants competed more for brand and cost efficiency, and exchanged sales volume for market, but how to survive and live a long time has become a problem that these enterprises have faced. always encountered.

However, characteristics of "fragmentation" and "decentralization" of entire security market determine that it is difficult to form a sustainable model in which all or a few dominate in field of intellectual security. With constant expansion and decline of smart security industry and scenarios, commercial implementation of smart security opens up new opportunities.

In field of public safety, "Project Snow Bright" proposed in 2016 has generated a market scale of tens of billions of dollars. This massive public safety prevention and control project to promote construction of a safe village has directly brought down safety market to county and village level, and has become a "pie" for integrators with rich products and strong engineering capabilities. As for those AI startups that cannot compete directly, they can take on role of a technology or server provider, take on some of engineering construction, and get a share of profits.

In B-end segment, intelligent security penetrates into segmented segments that are inaccessible to many industries. For example, at a construction site, original camera mayOnly play role of anti-theft surveillance, now by installing cameras on construction tower cranes and mechanical equipment of vehicles, it is possible to effectively monitor construction progress, personnel performance and security measures. Similarly, in ports, mining areas, for B-end industries such as fire protection, environmental protection and logistics production, location of AI cameras can greatly reduce workload of manual operations and on-site inspections, and security system can monitor and real-time analysis, and report hazards. Both AI algorithm vendors and application integrators can create their own professional capabilities in a wide range of new scenarios.

In some industries with complex projects and low margins, such as community intellectual security, situation is changing from an embarrassing situation where large manufacturers are unwilling to do this, but small producers cannot, to a situation where multiple parties are actively involved. . Some large developers have been trying to implement overall AI community security planning since beginning of community planning, and Ali and Tencent are also trying to connect to community services through their own software portals and build online AI communities in background. The possibilities of artificial intelligence; and many artificial intelligence companies are also implementing a full set of intelligent security community solutions that combine face recognition, video structuring, big data and other technologies with community scene applications to achieve more separated scene functions.

Sink, avoiding transition from traffic and public safety, which are crowded with giants on track, to more fragmented, but significant divisions in market, to gain living space; Take on a professional technical position in a split stage or provide a professional landing plan and become an industry giant's partner and provider of professional technology, products and services. These two directions can be a new choice for small and medium-sized enterprises to open new paths in smart security market.

In different ecological niches, find a place with rich aquatic plants

"In next ten years, find right position, and there will be big fish."

This is actually viewpoint of a security industry expert who spoke at China 2019 Security Industry Empowerment Conference last year. The main logic of this point of view is: how to redesign your own products and systems in terms of user scenarios and find your own new positioning in terms of expansion is possiblethe security industry with help of artificial intelligence, 5G, IoT and other new technologies?

For development logic of intelligent security industry, there is also expansion of technology capabilities, customer needs, and scene applications. The key is how to accurately translate customer needs into practical scene applications. Actually get out.

This logic is same for leading security companies as it is for many small and medium security companies.

The biggest difference is that leading companies can independently complete entire process of deploying underlying platform, network, hardware, and application algorithms to get an intelligent security ecosystem for these high-profit projects and customers.

Small and medium-sized security companies can focus on providing smaller O&M solutions and services in selected industries so that they can receive orders from clients or institutions in small and medium-sized industries such as just community, shopping mall or mines etc.

In addition, small and medium-sized smart security companies can also join giant company's smart security ecosystem and become industry chain partners through platform integration. Unique advantages in size and other aspects to look for best development opportunities.

Of course, for a large number of small and medium-sized companies involved in intelligent security systems, compared with proven and well-established security manufacturers and cross-border giants, small and medium-sized enterprises have to face big challenges.

From technology to stage implementation, these smart security companies must address following challenges. For example, do a good job of translating knowledge about customer needs to avoid project failure due to differences between implementation of technical solutions and customer needs; control cost of technical solutions, understand affordability of customer prices, and avoid seamless exchange of solutions that cannot cover costs during cooperation; The most basic principle of security system is data security. If customer data is lost, loss will be irretrievable.

The above is a major consideration for many security customers when choosing intelligent security solutions and service providers, and for many small and medium-sized enterprises to win customer trust.

But in any case, second half of intellectual security industry has arrived. Whether it is traditional mass public safety scenarios or new industry segmentation scenarios opened up by new technologies, it will provide an extremely rich ecological niche for different people. The number of players in differentth positions determines your place of survival.

Of course, it's also important to remember that second half is also start of playoffs. Once heroes grow up and seats get smaller, those members who can't take their seats will also have a hard time avoiding being squeezed, eliminated, or even swallowed.

However, race for smart security systems has only just begun. In wake of industrial internet and urban intelligence driven by new technological changes such as artificial intelligence, 5G and IoT, not only is it worth redesigning all traditional security scenarios, but a new wave of smart security will be born in To G and For B markets. scenarios bringing market trillion dividends.

The start of 2020 may cause temporary stagnation in many industries, but for security industry, it has brought positive revelations. Modern cities still face many challenges to cope with major societal crises, and smart security will be ahead. made.

It can be foreseen that during this epidemic, those cross-border giants and AI companies that act quickly and offer effective solutions will become potential players in second half of intellectual security.