Collaborative robots can safely interact directly with humans. Unlike traditional industrial robots, collaborative robot expanded the connotation of the function of the robot "people" feature, make robots have a certain autonomous behavior and cooperative ability, but in the unstructured environment with a group of people, work is mixed and disorderly movement and mission, combining with the person's intelligence, dexterity and brawn, power, and the accuracy of the machine, make robot real partners, man-machine collaboration over similar to precision installation process, solve the limitations of traditional industrial robots

The development features of cooperative robot skills are mainly reflected in three aspects: ease of use, sensitivity and security.

1. Ease of use

Ease of use is the top priority in robot controller design, which involves specific skills including all kinds of graphical programming, process package, teaching method, instruction richness, secondary development platform, system extensibility (hard interface, soft protocol), offline programming software, etc. However, the drag-and-teach feature of the cooperative robot eliminates the messy programming and equipment of the traditional industrial robot, making it easier to operate. However, the current market of some collaborative robot products, the graphical interface they provide can only end some simple picking, picking and putting mission, there is still not strong usability problem. In order to improve the ease of use of collaborative robots, enterprises in the professional sector optimize the development of robot programming language, so that robots can get used to a messy working environment. Quick AGV+ robot arm adopts a patented graphical programming method, which simplifies the messy instructions into the form of combined buttons, and coordinates with the dragging to teach the end of the robot programming; With the continuous expansion of the use of collaborative robots, to improve the simplicity and ease of use of collaborative robots, to provide more simple and intuitive, more powerful programming methods will still be collaborative robot enterprises a long time to strive for the direction.

2. Sensitivity

From the perspective of software, most collaborative robot systems have the ability to learn and evolve, making their products sensitive and habitual. They mainly adopt enhanced learning methods and genetic programming to complete multi-machine cooperative operations. However, the current multi-robot learning and evolution is still at a relatively low level of behavior, and its learning and evolution mission and environment are very simple, when it is faced with a more chaotic mission and environment, there are time delay evaluation and combinatorial explosion problems. At the same time, the distributed learning and evolution of multiple agents is obviously different from the traditional centralized learning and evolution methods, and more useful behavior optimization methods are yet to be found.

From the perspective of hardware, at present, most of the cooperative robot technology products are multi-joint robots with 6 degrees of freedom and above, with small dead weight and load, and relatively sensitive installation method and movement arrangement. It is suitable for the professions with higher requirements of flexibility, sensitivity and precision, such as electronics, medicine, precision instruments, etc., to meet the operational requirements of more industrial production and service occupations.

3. Safety

As a prerequisite for the completion of man-machine cooperation, only by ensuring the personal safety of the operators, can we better work together and improve the power of the enterprise. In order to effectively regulate the design and safety of Collaborative Robots, the world organization for standardization issued the specification ISO/TS15066 Robots and robotic devices -- Collaborative Robots in March 2016. As a supplement to the safety specification for industrial Robots supporting ISO10218, the specification further clarifies the design details and system safety skills of Collaborative Robots.

With the transformation and upgrading of manufacturing operations, the 3C industry, represented by computers, communications and electronic products of consumption, is booming. The product life cycle is getting shorter and shorter, and the personalized needs of users are getting higher and higher. Faced with the production line change time needs to be greatly shortened, the working space is small and so on, the traditional "either active or started" production installation form has been difficult to deal with, a large number of work needs and robot human cooperation to end. Therefore, in order to establish a sensitive and efficient production system, man-machine collaboration becomes a useful solution. Man-machine collaboration is not only a demand to solve practical problems, but also a long-term choice.