2017 International Conference on Robotics and Automation
Twice a year, the world’s leading robotics researchers gather to showcase their latest research. The 2017 IEEE International Conference on Robotics and Automation (ICRA), took place in Singapore. We reviewed most of the commercially relevant papers and attended many technical sessions.
Latest trends in robotics and automation research
During our visit, we found that the bulk of research funding and activity in robotics is increasingly being focused on two areas—autonomous vehicles, or what are better known as self-driving cars -- and unmanned aerial vehicles, commonly called drones. These two fields of robotics, in addition to other technologies related to their respective ecosystems, are undergoing the fastest development. This is occurring at the same time in which interest continues to wane in other robotics fields such as general consumer robots and humanoids. The technologies being developed are not only for the two types of aforementioned fields of robotics, rather there are myriad emerging areas that are transforming what we currently view as the applications for current self-driving and aeronautical technologies.
Korean robotics companies of interest
We review technologies relevant to existing and emerging business areas, and assess how they will affect businesses going forward. We then examine listed related Korean robotics companies and gauge their technological and business prowess to determine their feasibility as investments in the fast-evolving robotics landscape.
ICRA academia and industry research areas of interest
Vision, assistive, medical, and rehab robots, sensors, self-driving cars, and drones
Twice a year, the world’s leading robotics researchers gather to showcase their latest research. The 2017 IEEE International Conference on Robotics and Automation (ICRA), took place in Singapore. We have reviewed most of the commercially relevant papers and attended many technical sessions.
During our visit, we found that the bulk of research funding and activity in robotics is increasingly being channeled into two areas—autonomous vehicles, or what are better known as self-driving cars--and unmanned aerial vehicles ((UAVS)), more commonly called drones. These two fields of robotics, as well as ecosystem-related technologies, are undergoing the fastest development in robotics while other r&d areas such as general consumer robots and humanoids continue to suffer declining interest. The technologies being developed are not only for the two types of robot rather there are numerous emerging areas that are transforming what we currently view as the applications for current self-driving and aeronautical technologies.
At the conference, representatives from Korea included experts from the research departments of Seoul National University, KAIST, Postech, and many others. From the corporate research side were researchers from Naver Labs, Kakao Corp, and several startup robotics companies. We review and demonstrate technologies relevant to existing and emerging business and investment areas, and assess how they will affect the same and other businesses going forward. We then examine promising publicly traded Korean robotics related companies and assess their technological and business prowess to determine their feasibility as players in the rapidly evolving world of robotics. We look at, Naver (NHNCF), Hanwha Techwin, Hankuk Carbon, Sungwoo Hitech, CJ Korea Express, and Curexo.
Notably, China’s DJI was the most visible commercial company at ICRA and sponsored the DJI Mobile Manipulator Challenge as part of its search for international talent in robotics to develop ground-based robots. By this, DJI made clear its strategy to become a player in the comprehensive RIoT (Robotic Internet of Things) ecosystem. While we previously thought the company was shifting its focus to becoming a camera play, it is clear from their activities at the conference that they, like other players from the US, Japan, Korea, and Germany such as Amazon (AMZN), Google (GOOGL) (NASDAQ:GOOG), Apple (AAPL), Toyota (TM), Fanuc (OTCPK:FANUY), Samsung Electronics (OTC:SSNLF), Hanwha Techwin, Kuka (OTCPK:KUKAF), and others developing RIoT systems.
Setting sights on the RIoT crown
Key to the development of RIoT technologies from Amazon’s Echo Look to Softbank’s (OTCPK:SFTBF) humanoid line of robots to major companies’ growing fleets of autonomous vehicles and drones, sits machine vision, for it is with this core technology that path planning and data collection are based—central to mobile robotics.
Conventional vision sensors perceive the environment as a sequence of frames. Successive frames hold large amounts of redundant information, wasting memory access, power, computational power, disk space, RAM, and time. Furthermore, every frame puts an identical exposure time on every pixel, making it difficult to work with images with very dark and very bright areas. Event-based cameras are a paradigm shift in camera technology. Rather than using full frames, the camera captures a minimal event set caused by changes in intensity. Since the camera only transfers the changes, they can capture rapid movements in the scene or of the camera. Event-based cameras are attracting growing interest from robotics and computer vision researchers as they feature sensors that have three key advantages: 1) a latency in microseconds; 2) a rate of measurement that is almost one million times faster than regular cameras; and 3) a high dynamic range that is six orders of magnitude bigger than that of standard cameras. Event-based sensors create opportunities which standard cameras simply cannot match. Event-based cameras enable the design of algorithms that can track very small objects in the moonlight at night, the construction of drones that have the maneuverability of a housefly. The enable robots to localize and map in the most challenging lighting conditions and at incredible speeds. These sensors and are now being widely adopted in robotics. In 2016 they were made known to the public with the announcement by Intel (INTC) and Bosch of a US$15mn investment in Chronocam, a developer of event-based cameras, and Samsung Electronics announcing its use with the IBM TrueNorth processor to recognize human gestures. In Korea, we believe companies such as Samsung and Samsung Electro-Mechanics and companies in their ecosystems are developing the requisite hardware as well as the processing, learning, and control methods needed to create new markets for these sensors. Some of the current application areas for these new vision sensors are surveillance and ambient sensing, fast robotics, which includes mobile drones and fixed robots, factory automation, microscopy, motion analysis, hydrodynamics, sleep research and chronobiology, fluorescent imaging, and particle tracking.
Assistive, rehabilitation, and medical robots
Silver cloud from ambient computing and robotization
One of the biggest growth opportunities in assistive and rehabilitative robotics will be the ambient assisted living products in homes, through integration with smart technologies and increasing robotization. Assistive and rehabilitative robots must learn to adapt to the needs of the elderly and physically challenged. These robots must also have perceptual, cognitive, and bodily kinesthetic capabilities that are intuitive for users to interact with. Accordingly, companies and organizations that take a multidisciplinary approach to engineering and human- robot interface ((HRI)) research and develop the next generation of assistive and rehabilitative robots will have a decided market advantage. One of the greatest growth opportunities in this segment will be the ambient assisted-living products for the home, through the integration with smart technologies and robots.
We view Curexo’s recent acquisition of the robotics business from Hyundai Heavy Industries for KRW11.1bn as a good move for both parties as it enables the former to accelerate its drive into the global assistive and rehabilitative robot businesses and the latter to focus on its core competencies. The former Hyundai Robotics had been developing and deploying assistive and rehabilitative robots in some of Korea’s leading healthcare facilities. We believe the purchase of the unit will provide Curexo with a range of robots and robotics technologies that will give help it become a leading contender in these rapidly developing markets.
Ambient computing and robotics are giving rise to a field called Ambient Assisted Living ((AAL)), which aims to support caregiving, enhance the quality of life for the elderly and physically challenged, and to empower them to live safely and independently for longer in their own homes.
AAL is one of the key target applications of Ambient Intelligence ((AmI)) in healthcare and smart homes are the most widespread implementation. AmI is made up of the mechanisms that compose the infrastructure of smart environments, including robots, to provide environments that support and assist people.
More and more homes are being equipped with sensors for motion detection, the opening of doors and cabinets, and presence in bed. They are now mining the collected data to understand activity patterns. Pattern changes could be a signal of an accident that the AmI could address. A robot would then act as a first responder, complementing the AmI.
Surgical and other medical robots
With the global growth of the RIoT, medical robots are transforming from an idea of the future to a reality. Recently, medtech companies have been developing robots for the healthcare industry. Although Western countries are currently the largest market for medical robots, Asian governments and companies are aggressively investing in the development of medical robots.
Robotic surgery usually refers to minimally invasive surgeries (MIS) in which surgeons tele-operate robotic arms to operate on patients. Robotic surgery provides increased precision, smaller variations between surgeons, greater safety, fewer complications, improved patient outcomes, and faster recoveries. The biggest disadvantage to robotic surgery is the cost, which is usually more expensive than traditional forms of surgery for now. The entry by new players will change this situation.
US-based Intuitive Surgical’s (ISRG) da Vinci surgical robot is the leader in the global robotic surgery market. Cleared by the US FDA in 2000, it is used in a wide variety of surgeries. Hospitals the world over use the da Vinci and it continues to be adopted at an accelerating rate.
China’s Tianjin University developed an MIS robot system called the Micro Hand S. With positive results from tests conducted in 2014-2015, the Micro Hand S has the potential to be a safe and effective system for surgeries in the near future.
In 2012, Hong Kong was the first in Asia to purchase the Robotic Interactive Orthopedic (RIO) system from US company MAKO Surgical Corporation. The RIO is typically used for partial knee replacement surgeries, but can be used for hip, ligament, and spinal surgeries as well. The RIO lets the surgeon to guide the robotic arm during surgery, rather than operate it remotely on a console. Surgeons can perform more complicated surgeries with the RIO, and they can occasionally do it with less experience.
Relative to its population, Taiwan has the greatest number of surgical robots in Asia. Despite the many benefits of robotic surgery, hospital administrators in Taiwan are being challenged for using pricey robot-assisted surgeries in lieu of traditional surgeries.
In Japan, the government is working with universities and companies to develop a robotic surgery system that not only enables higher precision and accuracy, but will simultaneously monitor MRI readings and data from other devices. One of the latest robots to emerge from this collaboration EMARO is an air-powered endoscopic
robot that gives surgeons hands-free control and enhances visualization inside a patient’s body with three movable cameras. It is expected to be marketed to hospitals and universities in Japan in 2017 and be available overseas within the next few years.
Korea is also rapidly adapting surgical robots for healthcare. A training center was opened in 2008 at a Korean hospital for surgeons to practice the operating surgical robots. Inspired by the da Vinci system, Meere Company has been developing the Revo-i system since 2007. Revo-i will be used to treat cancer.
Koh Young Technologyis developing a neurosurgical robot for microscopic procedures on the brain and spinal cord. In 2007, Curexo co-developed Robodoc, an orthopedic surgical robot. Hyundai Heavy invested in Robodoc and has developed medical robots as well, including the Morning Walk for rehabilitation and walking. Curexo acquired the robotics business from Hyundai Heavy Industry in March 2017 for KRW11.8bn. We believe that Curexo’s focus on the fast-growing orthopedic surgical market with its US FDA-approved Robodoc and recent acquisition of Hyundai Robotics, with its line of tested rehabilitative robots makes it one of the most promising surgical robotics company in Korea.
Robotic surgery in India is currently in its infancy, unlike in China, Japan, and Korea. The Indian government purchased several surgical robots over the past two years for state hospitals. However, the technology and training are not yet easily accessible in India. That said, medtech companies and government hospitals are working to prove robotic surgery training to a minimum of 500 surgeons by 2020.
The already vast size of India’s growing population and wealth creates a unique opportunity for Korean medical robotics companies such as Curexo, Koh Young, and Meere as they already possess the technical knowhow and experience to meet the subcontinent’s burgeoning demand.
Other types of medical robots
Researchers continue to adapt robotic technologies from other industries for healthcare. One example, the automated guided vehicles ((AGVs)) used to transport heavy items can be used to carry objects or transport people in healthcare facilities. Panasonic’s (PCRFY) autonomous delivery robot, HOSPI, transports medicine, patient files, and lab specimens throughout hospitals. It is now undergoing trial runs in a Singapore hospital.
Singapore-based robot company Abacus Global Technology developed the AmaKris SR 2, a robot that diagnoses colon cancer and performs colonoscopies. This will lower costs, allow less-experienced doctors to conduct colonoscopies, and facilitate advanced telemedicine. Like Curexo, Abacus also seeks to offer a comprehensive medical robotics portfolio and builds service robots that transport patients and laboratory samples, as well as sensors that determine patient location and movement.
US-based Moog (MOG.A) showed several robots at ICRA including Simodont, a haptic-enabled dental-training robot. Simodont provides high-end dental simulation and training. It is a complete, proven training system for dental schools committed to greater data analysis and efficiency.
Cobots and sensor-based object manipulation
Recent developments in robotics and sensor technologies coupled with market forces are contributing to the rapid growth of the collaborative robot (cobot) business. In April 2017, Germany’s DHL (OTCPK:DPSGY) announced a partnership with US-based Locus Robotics. And it’s not just in the factory or the warehouse. Previously we mentioned how in the digital transformation of healthcare, there is a key role for the Internet of Things and for robotics, including cobots.
Locus Robotics expects a 50 percent increase of robots in hospitals for delivery purposes by 2019. And there is even more growth potential in the consumer space.
The idea of cobots nor the use of cobots is new. Yet, usage and developments are accelerating faster than many believe and cobots play a big role in the development of Industry 4.0 and the Industrial Internet of things.
How cobots differ from existing industrial robots
There are many of cases where cobots create value, primarily in operations that humans cannot or choose not
to fully automate. For example, the collaboration between DHL and Locus Robotics, the task is piece picking order
fulfilment. In general, cobots are typically tested and used for repetitive tasks whereby intelligent systems using
Collaborative aspect is key for the design, use, and evolution of cobots
The varied challenges in making robots work next to human staff in a secure manner have been solved and we are now at a tipping point whereby we only need to deploy the technologies, much as ICT was during the digital revolution. IDC forecasts that by 2018, 30 percent of all new deployments of robots will be smart cobots that operate three times faster and work with humans.
Cobots are usually mobile or can be easily moved, which is not the case with a traditional industrial robot. A cobot is typically smarter than a classic industrial robot. This does not mean that cobots are more complex and other industrial robots simple, what it means is that cobots are equipped with sensors, smart technologies, and systems which are linked with the Internet of Things or other systems. This is because cobots need to assist humans and these smart technologies and sensors are required as they are typically aware of humans, location, and context.
Because cobots are affordable, highly adaptable, and virtually plug-and-play, small and medium-sized enterprises ((SMEs)) are rapidly adopting them, and this segment will see tremendous growth in the next few years. Hanwha Techwin, with its growing line of commercial and military mobile robots, drones, and cobots, is positioned to capitalize on this market.
The bulk of service robots are collaborative by design, accordingly cobots are for more than simply SMEs or manufacturing. Offices, distribution centers homes, labs, warehouses, farms, and healthcare facilities are all looking to service robots to help them do their jobs more efficiently. The market is growing but uses for cobots already include packaging, assembly, machine tending, and material handling. Cobots also count, pick and place, and inspect. As cobots are increasingly deployed, new uses will be discovered.
Human-robot collaboration productivity gains will accelerate cobot adoption
The Massachusetts Institute of Technology conducted a human-machine study at a BMW factory and found that teams made of humans and robots working together to be more productive than teams made of either humans or robots alone. In addition, the cooperative process lowered human idle time by a tremendous 85%.
Sales of cobots currently account for less than 10 percent of the entire robot market but should soon undergo a strong growth period. Market acceptance of cobots and also drones will be significant drivers in the growth of non-industrial robots. This non-industrial growth will fragment and evolve into a variety of use cases for differing aspects of the service robotics marketplace but growth will be tremendous, exponential, and take place in the near future.
From a conservative view, cobot sales should grow roughly tenfold between 2015 and 2020, reaching over USD1bn from approximately USD95mn in 2014. Some analysts suggest more rapid growth: cobots will become the industry’s top seller in about two years, selling hundreds of thousands and with prices falling to the USD15,000-20,000 range. Market research firm goes so far as to forecast the global cobot market to grow at a CAGR of 50.88 percent to 2019.
Cobot evolution is accelerating
Increasing adoption of cobots by companies of all sizes
As robots migrate from behind fixed and caged locations to work next to humans, and as they move from industrial applications to assist and augment, companies will increasingly start seeing the evolving business landscape and they will have no choice but to jump in if they do not want to be left out the productivity revolution. Accordingly, companies around the world are acquiring or invest in robotics and robotic ventures to add to their catalog of products and services.
Teradyne’s (NYSE:TER) Universal Robots has a significant head start. Next year will likely see them continue to lead but FANUC, Kuka, ABB (ABB), and Hanwha Techwin, will begin to make inroads and experiment with different prices. By 2018, the competition will intensify as the number of cobots sold nears 15,000 units, with USD500mn in sales. It is still too early in the evolution of cobots for industry consolidation, but some systems will be preferred due to of their flexibility, ease of training, and support base.
To better see the potential for cobot sales we need to examine the size of the SME marketplace. There are six million companies around the world and they account for almost 70 percent of the world’s manufacturing. A few low-cost cobots would greatly improve the manufacturing process of the majority of these companies, and with returns that are easy to gauge. This means that cobots are en route to sell hundreds of thousands of units beginning in 2018.
Robotics and vehicular technologies for self-driving cars
Convergence of technologies and brainpower driving segment
The accelerating growth of autonomous car research is pushing the envelope for the state-of-the art vehicular technologies. This evolution is driving demand for robust planning algorithms, better sensing hardware, and superior control methods. At this point, robotics researchers have identified and addressed the most urgent issues for the development of the foundation of autonomous cars. This field is now being rapidly developed by the robotics community who is building more sophisticated applications such as intelligent and cooperative networks. State-of-the-art self-driving car technologies are now being developed from different approaches by experts in academia, industry, and government who are working together as a community to prepare the next level of an autonomous intelligent world.
IBM’s cognitive system for self-driving vehicle management
IBM (IBM: BUY) scientists developed a self-driving management system with the understanding of the human brain’s cognition and behavior generation. Their backgrounds as computational neuroscientists led them to devise a cognitive model and technique that uses sensors and AI to dynamically determine potential safety concerns and control whether self-driving vehicles are operated autonomously or by relinquishing control to a human driver. According to IBM, automobiles are evolving from a mode of transport to a moving data center outfitted with sensors and computers that capture information.
Naver and Sungwoo Hitech are Korean companies to watch
Naver Labs is aggressively developing self-driving and other robotics technologies for the Korean market. We believe that the company holds promise to be a leading player in the Korean market for self-driving technology and will monitor its development in the area closely. The company is planning to invest USD430mn in IoT, autonomous driving, and robotics. In addition, plans to offer a self-driving car rental service.
Diversifying its revenue sources from merely advertising into these new areas is a step in the right direction for the company, in our view. The company appears to be taking a page from Google’s playbook to boost its valuation.
We are even more enthusiastic about the recent move by Sungwoo Hitech to partner with Canada’s Phantom Intelligence-- a company that specializes in lidar. Phantom has provided LiDAR evaluation units to Korean automotive component manufacturers for the past two years.
The partnership will not only bring to bear Phantom’s development capability but will open the door for Sungwoo to move and interface with greater opportunities of North America and the EU thanks to Canada’s extensive and aggressive FTA stance. Sungwoo will also directly benefit from Ottawa’s commitment to develop AI and robotics in the form of tax incentives, relaxed immigration, and unrivaled industry-university research arrangements.
Drones growing more prevalent in robotics research
Deep learning, high-speed flying robots coming
Deep learning in autonomous drones is a fast-growing research field in industry and academia. Many companies are looking for ways to reduce the costs and risks associated with using pricey human labor for inspection and other work. By developing autonomous, high-performance drones, companies can take the human out of the loop, securing their data and lowering costs.
For Korea, we see the JV for development of FE-Panther VTOL UAV, a VTOL drone between Israeli Air Industries and Hankuk Carbon as a potentially large opportunity. The former is one of the world’s leading drone designers and manufacturer and the development partnership with Hankuk is a great opportunity for the cash-rich Korean materials company to enter a fast-growing market. The aerospace market’s estimated 5% annual growth makes it an industry for Hankuk to enter. Hankuk should be able to expand its business with its composite material production capability coupled with IAI’s extensive knowledge of UAVs development and manufacturing, the JV has the potential to become a globally competitive VTOL UAV company. The UAV is powered by lithium polymer batteries and a gasoline engine, and can fly for eight hours with a 6kg payload. It has a 130km range and can reach speeds of100km/h, and has automatic VTOL capability.
Recently, Hankuk/IAI announced that they may add shipborne take-off and landing capabilities to their FE- Panther VTOL UAV, which would be available by the end of 2018. This would be done by the JV before end of 2018. Based on Teal Group forecasts, the companies estimate the potential market for the VTOL naval UAVs at USD4.3bn from 2015 to 2024.
Hankuk and IAI are also co-developing a hybrid propulsion system for VTOL UAVs funded by the Korea-Israel Industrial R&D Foundation.
Disclosure: I am/we are long TM, IBM,FANUY.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.
Source : https://seekingalpha.com/article/4084952-robotics-outlook-icra-17-insightsThanks you for read my article Robotics Outlook From ICRA 17 Insights