Vision Dialogue ·Exhibitor|Mr. Kevin Jiang, Co-founder of INDEMIND

VisionChina 2021 (Beijing) has come to a successful end, with high quality forum presentations from industry experts attracting many visitors to share this machine vision event.

On the show floor, we had the opportunity to interview Mr. Jiang Wen, the co-founder of INDEMIND. Mr. Jiang Wen gave a detailed introduction on the new robots and solutions he brought to the show, let's take a look!

1、What are the latest technologies and products that you have brought to VisionChina (Beijing) 2021? What are the highlights of your main products?

We have brought a complete set of AI solutions for robots to this exhibition. The first thing we would like to introduce is the INDEMIND OS robotics base operating system, which actually contains the sensor data processing capability, AI algorithm fusion and business logic execution related AI logic needed for the robotics base layer. With this integrated OS system, the customer is saving the process of developing the core robot technology when building the robot, which is equivalent to having the main AI capabilities of the robot.

We are also bringing two universal robot platforms, our RBN100, which is aimed at robots around the 100kg class, and the RBN10, which is aimed at around the 10kg class. These two universal robot platforms feature a complete reference design for the robot chassis structure, as well as a complete design package for hardware and sensors. Once the customer has the reference design and fitted their own business unit on it, they can go about building a complete robot product at a very fast pace. Both robots are indoor scenario products, with some differences in the cost of sensors based on weight, and their core technology is a navigation technology based on stereo vision. The RBN100, a commercial model, is 80% cheaper than those currently on the market, and the RBN10 home model has a cost reduction of approximately 50%, which is the main product highlight of both robots.

2. What are the requirements for these two robots in terms of application scenarios? What are the options for the industry?

The RBN100 concept is a commercial robot for all scenarios, it achieves a 99% coverage of indoor commercial applications, including escalators, glass doors, glass coffee tables and other difficult problems that it can solve. So in terms of application scenarios, there is no limit to the scope. In terms of specific industries, because commercial robots are still in an early stage of implementation, for example, commercial distribution, logistics-related, leading, security and so on are all applicable, so it actually has no such industry boundaries.

The RBN10 has achieved 100% domestic applicability compared to the RBN100, and can cover any scenario we can imagine. This is all based on its recognition interaction, recognition interaction system has dozens of categories of hundreds of subdivision objects, basically household conventional objects can be recognized, also support the common household scene interaction function.

3. What is the biggest difference between the robots developed by your company and those that appeared on the market more than ten years ago?

I'll start by briefly explaining the history of floor cleaning robots, which are actually divided into three main generations. The first generation of sweepers appeared about 20 years ago and was a collision type, which used a randomised algorithm to make the robot collide around the house to achieve a convenient effect, it had two problems, the first was low efficiency and the second was that it was easy to get stuck in a dead end and difficult to return. The second generation of robots is the LIDAR scanning robots that appeared after 15 years. This generation of robots has solved the problem of work efficiency by building maps and can be recovered stably, and the experience of using them has been greatly improved. However, due to the single-line LIDAR, its obstacle avoidance capability is still inadequate. The robot we produce is the third generation, a pure vision robot based on visual 3D mapping. We have integrated navigation and obstacle avoidance technology all in one into one set of sensors, which has greatly reduced production costs. We are the only one that can do decision making interaction, which can directly recognise voice information and do it accurately, which means that the most intuitive experience of the robot is that it looks smarter. It's also a feature of our third generation products.

4. You are the only company to bring a vision robot to the conference, what are your thoughts on the future of machine vision?

Machine vision is the core of the entire natural navigation technology, because stereo vision is a bionic technology, it can theoretically solve all problems, but the current degree of excavation of vision is not enough, so it can not solve all problems. We can see that, including Tesla, including Baidu, in fact, their self-driving technology has been in the direction of vision to turn, enough to foresee the development prospects of machine vision. The logic of the application of stereo vision in the robotics industry is to develop algorithms, computing power and computing platforms, and then let vision take on more of the role of the most central sensor. It can solve the comprehensive information acquisition problem that a pure measurement sensor like LiDAR cannot solve, while further reducing costs. As the maturity of machine vision increases, vision will squeeze all other sensors and become the only source of sensors for the future robotics industry.

5、From the perspective of the current industrial environment, where will the company focus its business next?

We will soon be launching our RBN1 (one kilogram scale) and RBN500 (500 kilogram scale) robot platform products. Down to micro robots, such as audio robots and toy robots, and up to low-speed autonomous driving and industrial forklifts, these are the areas we cover. We won't touch road autopilot above L4 in the short term, but for limited scenarios of low-speed autopilot for large robots, all the way up to small micro-robots, these are the areas we specialise in. So we will expand our technology up and down to achieve 100% coverage of the robotics industry as far as possible.

6、Was your company satisfied with the quality and quantity of visitors to the show? Did you meet some important customers or potential customers? What are your expectations for next year's show?

We were satisfied with the show, we met some top global companies who approached us to discuss and cooperate with us, and we also worked more closely with customers from the machine vision industry. Our robotics business is spread across the country and our R&D and core technologies are concentrated in Beijing. For a professional conference, VisionChina is the ceiling of the machine vision industry in Beijing and we are very happy to attend.

VisionChina All rights reserved Jing ICP B. No. 16034613