The first report on the development of collaborative robot technology in China was released, and the four cutting-edge trends were overlooked

On June 22, Lide Think Tank, a high-end think tank in the robotics industry, and Jieka Robot, a global leader in flexible intelligent robots, jointly released the "2022 China Collaborative Robot Technology Development Report" (hereinafter referred to as the "Report"). The report analyzes the cutting-edge technologies of collaborative robots, and discusses the development trends of collaborative robot intelligent perception, autonomous cognition, human-computer interaction, collision detection and other technologies.

A number of industry experts from Shanghai Jiao Tong University, Beijing Institute of Technology, Shanghai Robotics Industry Technology Research Institute, Lead Robotics Research Institute and Jieka Robotics attended the online conference to explore the key technologies and innovative applications of collaborative robots.

Intellisense · Multidisciplinary integration

Perception is fundamental to the interactions between cobots and people, cobots and their environment, and cobots. As far as perception technology is concerned, in addition to multi-sensor information fusion, collaborative robots are increasingly showing a trend of deep cross-integration with neuroscience, biotechnology, artificial intelligence, cognitive science, and network big data technology.

Future research directions are theories and methods of active perception and natural interaction, with the addition of more sensors to enable cobots to understand human commands (through sounds, gestures, and graphics). Based on the research on the active acquisition, learning and reasoning methods of knowledge in complex and dynamic environments, the understanding and prediction theory of visual cognition and active behavior intention based on dynamic environment, the autonomous learning and knowledge multiplication methods of collaborative robots, and the situational awareness and natural interaction methods of multimodal human-computer collaboration, the mutual intention understanding, information exchange, and natural and harmonious emotional interaction between collaborative robots and humans are realized.

Autonomous cognition · Dexterous operation in complex environments

A high degree of intelligence is a major common technical requirement for a new generation of cobots. At present, collaborative robots are becoming more and more widely used, and they need to be more intelligent in terms of complex operation capabilities, adaptive and reconfigurable assembly capabilities, awareness of unstructured environments, and collaboration capabilities with people.

Collaborative robots are used in the commercial service market, and the main challenges include natural interaction, human-machine safety, environmental adaptation, complex and dexterous operations, etc., and intelligence is a feasible technical way to meet the challenges.

Collaborative robots work in complex and harsh industrial environments, requiring more intelligent environment perception and adaptability, human-machine collaborative operation capabilities, and exception handling capabilities. Advanced cognitive algorithms are the main means to achieve a high degree of intelligence in cobots.

Future research direction is autonomous cognition of complex environments. In-depth research on autonomous control for complex environments and complex tasks, identify the dynamic behavior of collaborative robots and the relationship and influence law between intelligent control and the environment, and realize the dexterous operation and autonomous control of collaborative robots.

Human-computer interaction · Intelligent robot operation

With the rapid development of human-computer interaction technology and the gradual deepening of research in the fields of ergonomics and user experience, some existing human-computer interaction systems are no longer limited to simple device input or teaching operations such as mice and keyboards, and some can complete the input of work programs through voice recognition, fingerprint recognition and even iris recognition.

People have higher and higher requirements for the usability and ease of use of collaborative robots, and the application of gesture understanding for human-computer interaction has become an important research content of human-computer interaction technology. Collaborative robots are one of the popular research directions of human-computer interaction technology through continuous gesture recognition, and have a wide range of application fields.

Gesture recognition is based on human hand movements, and in industrial production, the robot movement can be controlled through gestures, which can simplify the teaching process and operation process of the robot, which has great application value. Nowadays, the most prominent research methods in the field of gesture recognition include glove-based recognition and the use of visual information for recognition. The identification method using gloves is mainly to obtain and model the joint position and bending of the palm and fingers using optical fibers, etc. The vision-based recognition method refers to obtaining gesture visual images from the camera and processing them with a series of algorithms to recognize and obtain results.

Collision detection · Adaptive supple control

Collision detection generally includes sensor-based collision detection and sensorless collision detection.

Sensor-based collision detection can be based on electronic skin, joint torque sensor, or base, terminal six-dimensional sensor. In addition to this type of collision detection method, most of the other collision detection methods require the construction of a dynamic model of the cobot, the estimation of external forces, or the construction of an observer. When an external force or observer detects interference, a collision occurs.

The first report on the development of collaborative robot technology in China was released, and the four cutting-edge trends were overlooked

Based on sensorless collision detection, firstly, the detection of the single-joint motor current when the person and the cobot are in contact with the cobot is used to determine whether the cobot is in soft contact or collision contact with the human. Secondly, according to the actual value of the current obtained by the current motor feedback and the ideal value in the calculation of the dynamic model, the fusion algorithm is designed to generate a time-varying power collision detection threshold, identify the accident collision or the conscious soft contact of humans, and solve the problem of safe collision after contact between humans and collaborative robots.

The first report on the development of collaborative robot technology in China was released, and the four cutting-edge trends were overlooked

Sections viewed

Contact

working hours