An exclusive interview with Cai Yingpeng, CEO of Inspire Robots: Focus on Micro Linear Servo Actuato
Time:2019-09-16 Source:

Before the establishment of Inspire Robots, Cai Yingpeng has been developed in the field of robotics for nearly 10 years. At the beginning, he mainly worked as ODM (original design manufacturer) for the whole machine. In the process of system integration, he found the gap and opportunity of domestic core components, so he established Inspire Robots, focusing on the development of core components products. Micro Linear Servo Actuator and bionic hands are the core products.

At present, the development of Robots in China is entering a period of rapid growth. According to the research report "China Robot Market Forecast, 2018-2022" recently released by IDC Global Robotics Research, The Chinese robot market continues to grow at a high speed, with a compound annual growth rate (CAGR) of 26.9% from 2017 to 2022. In 2022, the size of China’s robot market will reach us $80.52 billion, accounting for 38.3% of the global total.

However, the lack of core components of domestic robots is particularly serious, and it has been subject to international giants such as the United States, Japan and Europe for a long time. Relying on imports makes the cost of domestic robots relatively high, which makes it difficult to develop the whole machine industry and the integrated application of downstream. The localization of core components is an inevitable stage in the robot market and a key field supported by the state. The integration of core components such as servo motor, reducer and controller may be a new trend in the future development of robots.

Founded in May 2016, Beijing Inspire Robots Technology Co., Ltd. is positioned as a supplier of robot core components and intelligent operation solutions. Company's core team from BUAA Robotics Institute, the company's core product Micro Linear Servo Actuator and The Dexterous Hands has been successfully applied to the robot, artificial limbs, medical and industrial areas such as, the cumulative application for dozens of domestic and international invention patent, and obtains the “National High Technology Enterprise”, “Zhongguancun Chuying Rencai Company Award” and “Zhongguancun golden seed Company Award”.

Before the establishment of Inspire Robots, Cai Yingpeng has been developed in the field of robotics for nearly 10 years. At the beginning, he mainly worked as ODM (original design manufacturer) for the whole machine. In the process of system integration, he found the gap and opportunity of domestic core components, so he established Inspire Robots, focusing on the development of core components products. Micro Linear Servo Actuator and bionic hands are the core products.

What is the underlying logic of temporal growth? What are the difficulties and characteristics of its core product servo drive? What is the current status of the robot industry? With these problems, Titanium Media conducted an in-depth interview on Cai Yingpeng.

The following is an interview between Ti Media and Tsai Yingpeng.

Titanium Media:

What kind of work did you do before Create Inspire Robots?

Cai Yingpeng: I started my career early. From 2007 to 2008, I worked in MEI Robocon robot Competition of USTB. In 2008, I went to BUAA Robotics Institute for my master's degree. In 2009, I set up a robot studio in my dormitory with my roommates and started ODM (original Design Manufacturer) project. In 2012, I registered my first company, which is also responsible for the development of various customized robot systems, such as substation power inspection robot, teaching robot, service robot, military and other special application robot, etc. By the end of 2015, it was decided to make the transition.

Titanium Media: What considerations led you to decide on transition?

Cai Yingpeng: ODM is based on projects, which cannot be expanded and has limited development space. Of course, it has also accumulated a lot of project experience and industry resources. It is also in the process of system integration that I found that the micro and small integrated linear servo system is a blank market in China and can be applied to many places, which is a good entry point. The company's transformation from ODM to its own products and brands also brought some personal changes to me. In addition to business and technical connection with the project side, I also worked on the front-line technology development. Since the establishment of The time, I gradually withdrew from the technical level, and focused more on the strategy and management of the company.

Titanium Media: On a technical level, what are your areas of expertise?

Cai Yingpeng: Electromechanical control system. The earliest is to do wheeled mobile robots, including motor driven, multi-sensor fusion positioning, navigation, planning algorithm; Later, I did distribute control and intelligent decision system for swarm robots, and motion control algorithm for joint robots.

Titanium Media: What is the advantage of Inspire Robots and Micro Linear Servo Actuator?

Cai Yingpeng: The Micro Linear Servo Actuator integrated with a motor, reducer, sensor and closed-loop control system. As the core component of the robot, Micro Linear Servo Actuator is characterized by small size (total length of the shell is about 60mm, weight is about 20g), high precision (the accuracy of repeated fingertip positioning is ±0.02mm) and high-power density (the maximum thrust is more than 100N).

Taking The Dexterous Hands as an example, it adopts an innovative linear drive design, and integrates six miniature linear servo drivers and six pressure sensors, which can accurately simulate human hands to achieve accurate grasping action, and then applies to service robots, prosthetics, education and scientific research and other fields.

Most manufacturers used the wire transmission scheme before, the accuracy cannot be maintained, the grip force is relatively small. On the other hand, it is also a problem that users are concerned about in the process of using, that is, whether the finger has a force sensor and whether it has a force control function. To take a simple example, when you grab something light you can use a relatively light force, and when you grab something heavy you can invoke a large force. As for the layout of the force sensor, if the sensor is attached to the surface of the fingertip, the cost is relatively high, and it may not be able to detect all the force sensors; By adding a force sensor at the end of the linear actuator, no matter where the finger receives the force, it will be transmitted back through the connecting rod mechanism.

Titanium Media: What are the difficulties encountered in manufacturing micro linear servo drivers?

Cai Yingpeng: I have encountered many difficulties along the way. Two representative examples are as follows:

One is how to process the reducer after the design is completed. Because the overall size of Micro Linear Servo Actuator is very small, the outer diameter of the planetary reducer is less than 9mm, and the gear modulus is less than 0.2. We spent almost one and a half years, after countless development and trial production, finally solved a series of technical problems such as gear material, batch manufacturing process, assembly process, quality inspection method. Gear quality inspection is more interesting, because the size is too small to be measured directly with instruments. We use a microscope to take photos of the gear, and then through image processing methods to detect the diameter, concentricity, tooth shape and a series of characteristics of the gear. At first, gear manufacturing was unstable, and some of the gears were produced with burr edges or missing teeth. Using the principle of rotation invariance of gear, we rotated the picture of gear at a fixed Angle around the center of the circle of the gear, and superimposed the image after each rotation, and then compared it with any original picture. By this method, abnormal gears can be identified very quickly.

The second is screw processing technology, including screw mechanism and screw assembly. The screw and nut in the screw mechanism should be just right, with too tight resistance is too big, with too loose clearance is too big, and the material of the screw and nut is also very exquisite. As for the screws, we used M1.2 multiplied by more than 20 mm screws, which could not be bought on the market. We found many factories for trial production, and the rate of good products was also very low. We have to purchase CNC lathe, research screw manufacturing process, to solve these problems.

Each of our parts requires a very high processing accuracy, many also need to use special technology. At the beginning, we went to a factory with strength for production, but we were not willing to accept orders without large quantities. Even if we received quality and construction period, we might not be able to meet our requirements. We have to spend a sum of money to purchase some equipment, their own research and development of various parts of the processing technology, and small batch trial production. Although the initial investment is very large, not only to buy equipment, but also to study how to use and maintain the equipment, various materials and processing methods, fixture and so on, but now it seems that everything is worth it, most of the research and development results can be quickly verified in the internal, greatly shorten the product development cycle.

Titanium MediaWhat are the commercial prospects of The Dexterous Hnads?

Cai Yingpeng: Commercial level bionic hands mainly considers two dimensions: the first is practical, not limited to scientific research or performance, but can complete the work, and has relatively high reliability; The second is the consideration of price, so that users can afford to use, use value. A lot of people, when they think about dexterity, they focus on how many degrees of freedom there are. In terms of technology development, the more freedom, the more flexibility. However, high flexibility also brings high cost and low reliability, and increases the complexity of the control system. Therefore, the real demand of the market needs to be considered. Currently, 6 degrees of freedom and 12 joints can be satisfied in most application scenarios.

Titanium Media: What is the pricing strategy for Micro Linear Servo Actuator? What is the level of shipments? How about sales compared with last year?

Cai Yingpeng: At present, we may be the only one on the market for supply. According to different performance parameters and functions, the retail price of Micro Linear Servo Actuator ranges from more than 800 CNY to above 2000 CNY, and the price with force sensor will be higher. We are currently the only supplier of micro linear servo drives in the market, and by the end of August this year, we have completed nearly 10 times the total shipments of last year.

Titanium Media: Does Inspire Robots is in a monopoly position in market of Micro Linear Servo Actuator

Cai Yingpeng: There is no monopoly. It will take some time for the market to accept this new product. At present, more than 100 customers are using our micro linear servo drive products, and most of them are still in the trial or small batch test stage.

Titanium Media: What industries do most buyers come from? Will it focus on one industry?

Cai Yingpeng: According to the current data, the downstream is still relatively scattered. Although we are positioned as a robot company, in fact, our core components are not limited to the robot industry, but also to the medical, industrial and consumer markets. The volume cycle of core components is not very fast. It takes about one and a half years for our customers to form small batches from the beginning of contact with products. Our customers are distributed in various industries, and the distribution of purchase volume is relatively average, with overall upward growth, and some industrial customers put forward the demand of bulk ODM.

Titanium Media: Now in the industrial Internet eigenvalue overhaul as a popular, do you have other aspects of action in the near future, such as in the product layout level or joint friends to build ecological level?

Cai Yingpeng: I think we still need to focus, especially for startups. First of all, we should achieve sufficient advantages in the segmentation scene and then consider expanding. Polishing the product is the first prerequisite. We also have a lot of customers asking for more than core components, such as complete systems. In order to meet the needs of customers, we will carry out ecological cooperation with the advantageous units in the industry to create a complete set of solutions together.

Titanium Media: What do you think of the current robotics industry as a whole? What are your optimistic directions for the future?

Cai Yingpeng: There is a gap between the overall development of the industry and the investment of social capital, and even far from the expectation of capital. Robotics is essentially a manufacturing industry. The growth rate will not be fast, but the growth space is large enough. As long as we keep polishing our products and market, there will be opportunities. Our company positioning is the core components and intelligent operation, core components is the foundation, intelligent operation is the future. The biggest difference between robots and other intelligent hardware is that robots not only have intelligence, but also have operation ability.

Titanium Media: In your years of working with robotics, have you encountered anything interesting and learned something from it?

Cai Yingpeng: Back in 2010, Nokia launched the N8 phone as a turnaround hope and the last gasp of Symbian3. A domestic Internet company commissioned us (still in the studio of the dormitory at that time) to customize a mobile robot for N8 mobile phone, which was used to shoot N8 mobile phone advertising. At that time, we made a deformed robot according to the requirements of the Internet company: In the shrunken state, like a big goose egg, the mobile phone is placed inside the egg as the brain of the robot, and the robot can flash lights to the rhythm of the music played by the phone. Unfolded is a quadruped robot that can move around on a flat surface and even jump in place. The project took less than two months from communication to completion, barely catching up with the time limit of advertising shooting, and finished the advertising shooting together with a popular singer. In the end, the N8 failed to save Nokia from more innovative Rivals like Apple and other Android phone makers. Continuous innovation ability is the driving force for the survival and development of enterprises, especially for start-ups.