Brain-computer interface technology has made breakthroughs from science fiction to reality in the past half century. At present, the research and development of brain computer interface has been continuously explored in bionics, medical diagnosis and intervention, consumer electronics, industry, education, virtual reality and other fields. Speaking at the 2023 Zhongguancun Forum – Brain-Computer Interface Innovation and Development Sub-Forum, Wu Xia, a professor at Beijing Normal University, said that non-invasive brain-computer interface technology is very suitable for industrial safety monitoring and is expected to improve industrial production efficiency in the future.

Industrial safety has equipment safety, personnel safety, production safety three needs. In terms of personnel safety and production safety, the monitoring of workers’ physiological and mental states mainly focuses on vital signs, fatigue state and consciousness state, while the monitoring of production environment and state safety focuses on cognitive load, attention level, working memory and emotional state.

According to Wu Xia, from the perspective of current technical means, pre-service assessment and safety awareness education are the main aspects to ensure personnel safety, and the detection equipment used is only simple physiological state detection equipment such as bracelets. In terms of ensuring production safety, environmental monitoring is mainly carried out through regular training, job rotation and rest, or simple equipment such as sensors.

“At present, it is difficult for industrial safety detection methods to meet the needs of industrial safety monitoring. We hope to introduce some new technologies. Non-invasive brain-computer interface technology is very suitable for industrial safety monitoring, but there are three difficulties to overcome.” “Wu Xia said.

The first difficulty is to ensure the quality of high-precision brain telecommunication signal from the source of hardware design. For example, the full band 24 is used as the ADC EEG acquisition circuit to achieve more than 320,000 times of common-mode suppression capability, and carry out fine power processing, accurate ground plane processing and proper configuration of peripheral circuits in the control of noise level.

What is an easy-to-use brain-computer interface system? According to Wu Xia, firstly, industrial requirements such as waterproof and dustproof, low power consumption and anti-interference dry electrode should be met, and strict biocompatibility experiments and refined ergonomic design should be carried out in terms of human perception, and the overall weight should not exceed 130 grams.

The second difficulty is how to achieve highly stable and reliable algorithm output results in complex industrial scenarios. “The solution is to build a real-time adaptive EEG filter under complex conditions, and use as few channels of data as possible to further improve the robustness of the application algorithm and make the output more stable.” ‘she said.

It is understood that the creation of real-time adaptive EEG filter needs to effectively identify the environmental dust state, high humidity and other bad condition signals, and can be carried out in the personnel walking, climbing, loading, jumping and other motion state EEG acquisition, in addition to the need to support more than 1000 people synchronous acquisition, to constitute the basic human physiology and EEG parameters of the database for assistance.

The third difficulty is how to ensure the rapid and stable operation of the brain-computer interface system in various practical industrial environments. Wu Xia pointed out that in order to achieve this effect, first of all, the system stability should be stronger, the wireless signal transmission problem should be solved, and the monitoring and second-level response should be realized through the interoperable cross-platform application program interface. Secondly, it is necessary to build a real-time transmission analysis and processing system for thousands of people to realize multi-terminal and rapid standardization of subordinates. Finally, develop targeted solutions. For example, in the construction scenario of heavy gantry crane, a complete set of safety solutions is formed by combining brain-computer interface, wireless ranging, cargo type, and broadcast communication.

According to reports, the brain-computer interface based industrial personnel safety digital management system has been applied to tunnel construction, mining, foundation pit operation, power line patrol and other scenarios, providing nearly 400 risk reports, issued nearly 2000 risk warning and alarm information, and successfully captured nearly 30,000 risk data.

For example, in the tunnel construction scenario, the contact between large machinery and heavy materials and people in the process of transportation, hoisting and transfer will cause safety accidents. Therefore, the construction personnel can wear a helmet with a brain-computer interface device to monitor the brain status of the workers in real time. The workers can check the physiological characteristics of the construction personnel such as pulse, blood oxygen and so on through the real-time monitoring platform of employee status, and judge fatigue and concentration by analyzing EEG data. The scheme has been used to build tunnels in Guangzhou, Beijing and Chongqing.

“Looking into the future, in addition to being helpful to industry in terms of safety, brain-computer interface technology is also likely to improve machine efficiency through brain-controlled machinery, brain-controlled cars and other ways, and improve human efficiency through instrument operation, equipment driving and other skills training.” Wu Xia said.