In March 2026, the Compilation of Typical AI Application Scenarios in Airports, jointly compiled by the Key Laboratory of Smart Airport Theory and System of Civil Aviation and the Network Information Technology Application Branch of China Computer Users Association, was officially released. The "Mianyang Airport Iris System Application Solution" submitted by WuHan Homsh Technology Co., Ltd. was successfully selected and received an official inclusion certificate.
This marks the first time that iris recognition technology has been incorporated into the national civil aviation AI application case system as an independent technology category, signifying that Homsh's practical achievements in the smart airport field have received systematic recognition from authoritative industry institutions.
Inclusion Certificate
Inclusion Certificate for the Compilation of Typical AI Application Scenarios in Airports
An Authoritative Industry Panorama
The Compilation of Typical AI Application Scenarios in Airports is not an ordinary industry white paper. It was compiled by two authoritative institutions in the civil aviation system after months of extensive solicitation and strict screening from airports and technology suppliers across the country.
The book has more than 500 pages and includes dozens of typical application cases covering six major sections: Artificial Intelligence and Security, Artificial Intelligence and Operation, Artificial Intelligence and Services, Artificial Intelligence and Planning, Artificial Intelligence and Construction, and General AI Capability and Management Empowerment.
Participating enterprises include core operators and technology service providers in the civil aviation field, such as Guangzhou Baiyun International Airport, Hangzhou Xiaoshan International Airport, Shenzhen Airlines, and China Southern Airlines.
Homsh's solution was included in the "Intelligent Security Inspection and Biometric Recognition" section. From the table of contents, other cases in this section mostly focus on mainstream biometric recognition routes such as facial recognition and 3D visual recognition, while Homsh is the only enterprise with iris recognition as its core technology. This fact alone demonstrates that the irreplaceability of iris recognition in high-security scenarios is being recognized and accepted by the civil aviation industry.
The Invisible Line of Defense for Internal Airport Security
When the public talks about airport security, the first things that come to mind are passenger security check channels and entry-exit gates. However, for airport operators, another equally critical yet unseen line of defense is the access control of personnel in the airport's internal working areas.
The apron, control tower, equipment rooms, aircraft maintenance areas, and fuel storage areas — these areas are not open to passengers, yet they carry the core functions of airport operations. Unauthorized access to these areas can lead to far more serious consequences than security incidents in passenger channels.
For a long time, however, the internal working areas of most domestic airports have still relied on traditional access control methods, such as IC cards, employee badges, or passwords. These methods share a common flaw: they authenticate "objects", not "people". An employee badge can be borrowed, a password can be leaked, and an IC card can be copied. In places with the highest security requirements, this logic of "possession equals authorization" has fundamental loopholes.
The Mianyang Airport Iris Access Control System was precisely designed to solve this problem. The project started construction in August 2020, and was deployed and put into operation after three months. The core idea of the solution is clear and resolute: replace traditional credentials with non-replicable, non-transferable, and non-lost iris biometric features, and establish a high-security access control system that "recognizes people, not cards" in key internal areas of the airport.
System Architecture: The Complete Link from Registration to Access
The Mianyang Airport Iris Access Control System deploys iris access control terminals in key internal areas of the airport. Staff identity registration is completed through iris collection equipment, and iris feature templates are entered into the system. After that, staff can complete identity verification and access authorization for daily entry and exit with their irises alone.
Hierarchical Design Architecture
The system adopts a three-tier architecture of client layer, interface layer and service layer, and realizes optimal scheduling of computing resources through the high-speed iris matching engine and encoding engine.
Offline/Online Dual-Mode Operation
The front-end access control equipment supports both offline and online dual-mode operation: it synchronizes with the background in real time under normal network conditions; when the network is disconnected, a single device can support local matching for up to 2000 people, ensuring continuous control even during network outages.
Public Security Database Connection
The system is connected to the public security database, and synchronously generates alarm information for personnel with abnormal identities, realizing real-time early warning and precise control.
Technical Core: Triple Breakthroughs in Accuracy, Speed and Power Consumption
The technical requirements for an access control system may seem simple — "recognize the person and grant access". But in an airport environment, these six words come with extremely stringent technical indicators.
1.Accuracy: False Recognition Rate of One in a Billion
The Mianyang Airport solution adopts Homsh's self-developed PhaseIris iris recognition algorithm, using iris codes with a length of up to 64Kb, achieving a binocular false recognition rate as low as 10⁻⁹. At the same time, the system integrates a liveness detection function, which can effectively defend against spoofing attacks using iris photos, videos and other prostheses.
2.Speed: Access Completed in Two Seconds
The average iris collection time is about 0.7 seconds, and the recognition time is about 0.9 seconds. The entire access process can be completed within two seconds. The maximum matching speed of the system can reach 5.6 million feature values per second.
3.Power Consumption: 80% Energy Reduction Through Chip Design
In 2018, Homsh launched the Qianxin Q80, the world's first application-specific integrated circuit (ASIC) chip dedicated to iris recognition. The chip-based architecture reduces the power consumption of front-end equipment by approximately 80% compared with traditional pure software solutions, with an encoding speed of up to about 200 frames per second.
Compliance: Strict Alignment with National Standards
The Mianyang Airport Iris Access Control System complies with the technical specifications formulated by the Ministry of Public Security and international standardization organizations, including:
● GA/T 1286-2015 Iris Image Data
● ISO/IEC 19794-6 and other key standards
Standardized design brings not only compliance guarantees, but also replicability. The successful practice of Mianyang Airport can be extended to more airports and other high-security venues along a standardized path.
From a Single Gate to an Industry Signal
This selection conveys a noteworthy industry signal: in the civil aviation industry, which has the most stringent requirements for safety, iris recognition is evolving from an "optional solution" to the "preferred solution".
Iris texture is fully developed in infancy and remains unchanged for life, unaffected by external factors such as aging, wearing glasses, and makeup. The uniqueness of binocular irises even surpasses that of DNA identification.
Homsh has always promoted industrialization along the vertical integration path of "algorithm — chip — module — terminal — solution". This inclusion in the national civil aviation AI case compilation is an important milestone for this technical route to be verified in another key industry.
Conclusion
Airports are the first line of defense for civil aviation safety, and also one of the most technology-intensive scenarios in the smart transportation system.
Taking Mianyang Airport as a model, Homsh will continue to promote the large-scale application of iris recognition in civil aviation security, public transportation hubs, and more high-security scenarios —
To ensure that every critical gate opens only for the right person.
