Can I operate a drone by wearing a non-invasive brain computer interface device? Recently, the Haihe Laboratory of Brain Computer Science at Tianjin University and the School of Integrated Circuits at Tsinghua University jointly developed a new non-invasive brain computer interface system based on memristor neural morphology devices, and for the first time revealed the synergistic enhancement effect of EEG development and decoder evolution in brain computer interaction, successfully achieving efficient four degree of freedom control of drones by the human brain. The relevant results were recently published online in Nature Electronics. Brain computer interface can achieve direct information exchange between the brain and machines, promote the integration of biological intelligence and machine intelligence, and is recognized as the core technology of the new generation of human-computer interaction and human-machine hybrid intelligence. How to achieve "mutual learning" through information exchange between brain and computer, and promote the collaborative evolution of brain computer intelligence, is the key and difficult point to break through the bottleneck of brain computer performance. The research teams from Tianjin University and Tsinghua University have found that the non-stationary characteristics in EEG signals are not only derived from the background EEG variation traditionally believed, but also closely related to the task EEG evolution guided by closed-loop brain computer interaction. Based on this discovery, the team constructed a novel brain computer interface system using memristor neuromorphic devices and proposed a "dual loop brain computer co evolution framework". In the dual loop framework, the memristor decoder in the "machine learning" loop updates decoding parameters by adapting to fluctuations in EEG signals, and the task related EEG features in the "brain learning" loop continuously evolve positively under the guidance of the "decision feedback" loop. Compared to traditional pure digital hardware solutions, the new memristor scheme has increased the normalized decoding speed by more than 2 orders of magnitude (hundreds of times) and reduced energy consumption by less than 3 orders of magnitude (one thousandth), efficiently supporting the successful achievement of the four degree of freedom brain controlled drone mission objectives. In a continuous 6-hour long-term brain computer interaction experiment, the contribution ratio of the brain and decoder showed dynamic changes, and the final brain computer interface performance achieved an accuracy improvement of about 20%. This study first proposed the concept of brain computer co evolution and completed technical validation based on memristor neural morphology devices, providing important theoretical basis and technical support for future practical brain computer interface systems, and opening up new directions for the development of brain computer fusion intelligence Xu Minpeng, a selected member of the National High level Talent Program and a professor at the Haihe Laboratory of Brain Computer Science at Tianjin University, said. Our system plans to expand into more portable or wearable brain computer interface devices in the future, serving various intelligent human-computer interaction practical scenarios such as consumer level and medical level. It is reported that this research was carried out by a team from Tianjin University and Tsinghua University. The Tianjin University Brain Computer Haihe Laboratory team completed the collaborative evolution brain computer interface software system design and paradigm algorithm implementation, and the Tsinghua University School of Integrated Circuits team completed the collaborative evolution memristor neural morphology device hardware design and memristor algorithm design deployment. (New Society)