Problems To Be Solved

Although industrial robots have been widely used in the automotive manufacturing, metal manufacturing and other industries, industrial robots are generally not intelligent and not safe for human. The major problems we want to solve are as follows.

(1) Existing industrial robots are mainly concentrated on simple repetitive tasks such as laser welding, painting, handling, and loading and unloading. For manufacturing processes that require human judgment, most work are still done by human. With the current rising labor costs, the manufacturing industry is generally facing the problem of labor shortage.

(2) Most of the existing industrial robots do not have intelligence. They need to be programmed by highly-paid professionals, and they cannot perform frequent task changes according to product changes.

(3) Most robots need to be “closed in cages” to prevent harm to the human body. Humans and robots are completely separated and there is no cooperative relationship.

(4) Existing industrial robots are expensive, which prevents small and medium-sized enterprises from leveraging robotic solutions.

With the ever-changing of products and the increasing competition in the global market, global manufacturing companies are seeking smart and flexible robots. One of our tasks is to develop an intelligent robotic solution that can solve the above problems.

Solutions Being Developed

The solution we are developing is a soft collaborative robot with intelligence. This is a new generation of robots with fast learning, rapid deployment, high flexibility and adaptability. It has the ability to perform complex tasks with human. The solution has the following features:

(1) Robotic brain: The robotic brain will integrate the research results in machine vision, deep reinforcement learning, imitation learning, one-shot learning, meta-learning, intelligent autonomous system, and a new generation of control technology, making the robot efficiently learn the new skills needed to perform complex tasks and quickly adapt to new or complex environments (or tasks). The robotic brain enables robots to accurately handle more complex and broader tasks such as machining complex-shaped parts, picking complex boxes, assembling complex products, stacking/unstacking irregular objects, and manipulation of deformable objects, etc.

(2) Robot body: The robot body will integrate "soft" robot technology and adopt advanced human-machine collaborative architecture, featuring safety, easy operation, and openness. By closely working with human, it can perform complex tasks that general industrial robots cannot perform, providing manufacturers with the high-performance automation they need while maintaining affordability.

3) Adaptive technology: The robots will integrate the research results in intelligent autonomous systems, enabling robots to adapt themselves to complex environments. At the same time, the unprecedented range, shape, size, or weight of components can be processed to meet the speed and precision required by today's and tomorrow's manufacturing plants and warehouses. 

(4) Flexibility: The robots can quickly adapt to product changes, and quickly and flexibly deploy to various production lines, greatly reducing the time required for product manufacturing. For example, they can quickly switch tasks among parts processing, injection molding, welding, polishing, assembling, painting, packaging, labeling, and other work, even economically viable in small batch productions.