About indoor mobile robot, this technology in addition to the traditional AGV guidance technology (robots follow guide line), the main is a kind of amazon KIVA so let robot visual beacon solutions through the ground, robot follow track positioning and navigation for mobile or landmark, and is more advanced autonomous mobile solutions, The robot can realize autonomous positioning, obstacle avoidance and path planning in the process of walking. Compared with the previous two schemes, it can avoid laying guidance lines and pasting landmarks, which affect the environment and human activities.
AICRobo, a company based in Shenzhen, is a solution for autonomous robot mobility. The founder, She Yuanbo, engaged in software development after graduating from the bachelor's degree in 2009. In 2012, after accumulating several years of IT experience, he resigned from his job and entered the Institute of Machine Learning and Data Mining (LAMDA) of Artificial Intelligence Laboratory of Nanjing University under the tur of Professor Zhou Zhihua. Focus on the research direction of "vision-based robot navigation" study robot navigation related topics, do the development of robot autonomous mobile system.
In 2013, based on her familiarity with the service robot industry and her accumulation of knowledge and experience in mechanical electronics, software engineering and artificial intelligence, She predicted that the service robot industry would explode based on autonomous mobile technology, so she established AICRobo intelligent robot technology team.
In 2014, with the growth of market demand, I established a company with three or two like-minded friends to provide technical services of autonomous mobile robots for enterprises, which was the early prototype of the company.
She yuanbo, who graduated from NTU in July 2015, chose to continue her business and established AICRobo in Shenzhen in August. The company obtained angel round financing in May 2016. At present, there are 23 people in the team, including engineers in various fields including machinery, electronics, embedded, automation control, software engineering and artificial intelligence.
Positioning "indoor" robot autonomous mobility solutions, deep cultivation of vertical market
Unlike before, AICRobo not only provides intelligent technology software services, but also makes hardware chassis products. In addition, the company is to focus the focus on the "interior" usage scenarios, for security inspection guide, shopping, catering, health care, the robot products provide warehousing and other industries of autonomous mobile solutions, implementation environment building, autonomous navigation (real-time positioning, intelligent obstacle avoidance and path planning), intelligent follow, natural human computer interaction (voice interaction), and other major functions.
For the robot's "eye", AICRobo uses a combination of "lidar + ultrasound + vision" to complement each other, each responding to environmental perception under different conditions: Lidar is mainly used in the detection of plane obstacles, but because the infrared laser beam is easily affected by the material, and can not detect glass, mirror; Ultrasonic point ranging, in fact, is a small cone, can detect the infrared laser can not perceive the material; For vision, depth camera is used to detect obstacles in stereo area.
Due to the lack of guidance and positioning of guidance wires and landmarks, autonomous mobile robots need to have more complex functions to cope with the environment. According to Sheyuanbo, AICRobo developed a set of autonomous mobile system -- AICANS system, which not only meets the needs of autonomous mobile intelligent robots, but also based on the overall architecture of software and hardware integration. Providing interfaces for easy secondary development for different industry applications can reduce a lot of integration effort and shorten development time.
Measure yuan bo, moreover, to further explain the autonomous mobile robot platform with traditional AGV in technology: the difference between traditional AGV not have autonomous obstacle avoidance function, will only stop when meet obstacles, so need traffic control for scheduling, autonomous mobile robot can meet obstacles to planning route and bypass, the path planning and scheduling mechanism is more complicated. Moreover, the two sides differ greatly on many technical details.
"Indoor" robots were popularized much earlier than "outdoor" ones
"For autonomous mobility technology, outdoor solutions are much more complex than indoor solutions, including strong light, weather, and complex environment." She Yuanbo had long realized how deep the water was, so she did not plan to touch this market from the very beginning, such as the outdoor walking express robot. He thinks it is still too early to see such outdoor robots in real use, at least until 2020, and autonomous driving will be on the ground much longer.
At present, the indoor service robot products facing the public service field include the following categories: internal transport robot (such as hospital, office document delivery); Service transportation robot (such as food delivery, hotel); Shopping and browsing robots (mainly used in shopping malls, museums, real estate, etc.); Security inspection (airport, station), etc.
Measure yuan bo think, one of the earliest can be spread is storage, security inspection, medical applications such as robot, in addition, the autonomous mobile robot at the start stage, because many of the technical engineering, WanShanHua degree is not high, can't have good enough experience, but with the optimization and perfection of the autonomous mobile technology, This market will burst in the next 2-3 years.
In the robot autonomous mobile technology, although foreign leading domestic 2 years, domestic also have the opportunity to overtake
In fact, in foreign countries, the technology commercialization of autonomous mobile robots started in 2012, two years ahead of China (China started in 2014), and the product commercialization in foreign countries began in 2014. Compared with China, the product has a higher degree of perfection. She yuanbo said that foreign markets also have certain disadvantages, such as slow progress, high cost, and inability to provide customers with perfect customization, integration and after-sales service.
She yuanbo further said: "The future of intelligent robot must be closely combined with hardware and software agent, the ratio of base for software implementation even more, the core of the intelligent robot research and development of software and intellisense breakthrough is robot industry to realize our country to catch up with main direction, in addition, rapid can be customized is also a feature of the smart logistics solutions, "Different customers have different requirements for in-factory logistics business logic and application scenarios, so it is necessary to have this rapidly customized business logic driven software platform, loosely coupled with the base platform."
Based on the accumulation of engineering strength and experience, they know how to improve in the overall architecture
At present, AICRobo is developing the second generation of autonomous mobile solutions and solution-based products, which are expected to be completed in June this year. Meanwhile, AICRobo is also cooperating with universities, research institutes and customer enterprises to develop extended products. In addition, AICRobo now has more than ten customers, covering hotel service, welcome guests, food delivery, medical and other fields and applications.
In the future, AICRobo will continue to invest in the repair and improvement of algorithms, system optimization and other technologies, including improving the robustness of positioning, real-time obstacle avoidance and system reliability.
"In the process of technology into products, need manpower and energy of continuous investment, our advantage lies in the strength of the engineering and experience in mobile robot overall architecture, including algorithm, sensors, embedded development level, can improve the timeliness of response and under the overall architecture and can do it in the field of low cost. There are a lot of open source solutions out there, but it doesn't matter. The transition from open source to Productization takes a long time. After all, we know how to optimize the rest." She yuanbo said. (Author: Jes@Shenzhen Bay) ■
