With the substantial growth in venture capital (VC) transactions, the 'robot' area is rapidly emerging and is expected to see the trend of steep growth in revenue after 2020. Industry experts and expect that as more and more Chinese factories start Operation and the use of more automated systems, you may also see the market long-term stable growth and price cuts.
According to Dan Kara, director of research at ABI Research Robotics, industrial robots are expected to grow strongly after 2020 due to the rapid expansion in Asia and the emergence of more low-cost, high-quality systems, and he expects China to surpass the US in a few years , Becoming the largest industrial robot market.
Kara said in a recent RoboBusiness robot technology conference in Silicon Valley that robots have long been out of the 'precision drive system', both of which are made in Germany, Japan and Switzerland, and now China is catching up and The more systems use intelligent software, so no longer need those expensive drive systems.
ABI Research also expects global industrial robots to reach 500,000 units by 2020, up from 36 million units in the previous year, and China will be an important force in driving this growth curve and driving the market 2025 years ago to break through the sales of millions of units.
Kara said that venture operators are optimistic about the development potential of the robot market, from 2015 onwards can see VC in the field of robot investment transactions increased significantly.The amount of investment in the robot more than a dozen sub-areas, including software and The sensor occupies the largest part.
In addition, LiDar, which is widely used in the latest system, has been the most promising focus area for investors, and in fact the two companies, Quanergy and Velodyne, have received the highest investment amount of $ 90 million And $ 150 million.
Collaborative robots also understand teamwork
Collaborative robots and mobile robots are the hottest, and various mobile systems from small seats to large lifts are flooded throughout the manufacturing plant or warehouse.
Collaborative systems often use some form of machine learning software to learn how to work with other people, for example, Fanuc's latest Edge factory machine uses Preferred Networks software. Kara said: 'Only a few hours, the robot will be like As human beings quickly pick up and feed materials to other robots.
This type of machine learning is the latest stage in the development of robotic software, and the previous generation of robots is mainly learning from human behavior, and before that, human operators must program instructions in the software to guide the robot.
Kara predicts that collaborative robots will reach $ 1 billion in market size by 2020, with more than 50,000 systems sold each year, while unit prices will continue to decline. Mobile robots have a strategic advantage, so Amazon has Its storage to build their own robot, there are several new companies are also very attractive.
In the field of various segments, the wearable exoskeleton robot is also a market that now seems small but has great potential.Kara pointed out that the field of a dozen application cases, and have the opportunity to reach the scale of 2.6 million units per year, Of which more than half are agricultural applications.
As for consumer applications, he believes that consumer robots are not particularly useful in addition to popular vacuum cleaners and swimming pool clean robots, but added that voice assistants like Alexa are the most influential in the home.
Robot swept the retail industry
Fellow Robots Technology Thavidu Ranatunga (left) and software engineer Hadi Tabatabaee said that the industry has yet to find a good way to locate the indoor positioning technology.For example, their latest system using two light sensor and 100M pixel high-resolution camera, But like other mobile robots, it will still be lost.
They said that engineers have to use a variety of methods to integrate multiple sensor positioning data to improve reliability, but the current performance is already good enough, such as the San Francisco Bay Area's Lowe's and BevMo retail stores as early as 2014 Year began to use the robot.
Debugging is another big problem recently encountered in developing robots, and their current systems use x86 and Nvidia GPU computers to connect to the sensor via an Ethernet router to provide a more reliable connection than USB. , The complex interdependencies between the various subsystems, making the solution to software errors become more challenging.
Looking at the remote robot market
SRL International Senior Research Engineer John Marlow presents a prototype of the dismantling robot, the Taurus, which is built by the US Department of Homeland Security, which has been used by the FBI for the US Federal Bureau of Investigation (FBI) Distal demolition of bombs.
Taurus robotic robots use the same technology as Intuitive Surgical - Intuitive Surgical is a robotics company that was spin-off from SRI more than a decade ago. This technology is used to help surgeons make it easier to use a stereoscopic 3D interface To manipulate and control.
At present, systems like Taurus cost as much as $ 200,000, partly because they use custom tools to create only one robot at a time. SRI wants to find other manufacturing methods to reduce this type of system to 30,000 US dollars, and find new market applications, such as Chile's mining industry plans to use SRI technology, set up in the distal burst.
Robot Seconds Magic Square
ABB is one of the many companies that show machine arm with machine vision. The most interesting one shows that robots play magic squares.
Compiler: Susan Hong