What it is and the value it drives
A pioneer in the field and still an industry leader today, Japan has been populating its factories with robots since the 1970s. Not necessarily an intelligent humanoid, a robot is, in fact, any machine with three programmable axes that can execute a series of complex instructions automatically. Over the last centuries, they have been designed in a multitude of forms and functions.
Japanese companies like Mitsubishi, Omron, FANUC, and Yaskawa have helped make the country the number one industrial robot manufacturer, and currently Japan's second-largest firm, Keyence, is a robotics supplier, valued at $100Bn. However, in recent years Japan lost its long-standing record for the highest density of robots per employee to Singapore, South Korea, and Germany; moreover, with their strong focus on advanced software, several overseas players such as Universal Robots, Boston Dynamics (owned by the SoftBank Group), and Rethink Robotics are emerging as worthwhile contenders.
Robots can be divided into four classes, with different levels of commercial maturity.
Stand-alone robots. The most traditional form of industrial robot, these are fixed-place manipulators that can be programmed to perform tasks by operating in isolation, often with safety fences or equipment in place.
Intelligent Autonomous Guided Vehicles (AGV). AGVs are mobile robots that can navigate an environment independently and perform logistical tasks.
Collaborative robots (cobots). These are robots with built-in safety features that make them capable of working side by side with humans, or assist them in delicate, precise, or labor-intensive tasks, without the need for special safety equipment.
Microbots. These are miniature robots, often deployed in self-coordinating swarms, which can unintrusively operate in small spaces, including the human body. Out of the four classes, microbots are the class that has not yet reached commercial maturity.
Robots have so far found the widest adoption in manufacturing plants for the following key reasons:
Health and safety. Tasks requiring strenuous work, such as lifting heavy loads, or taking place in hazardous conditions, such as at high elevations or in proximity to toxic substances, often put workers at risk of health issues, serious injuries, or even death. While introducing semiautonomous machinery can pose risks of its own if uncontrolled, robots deployed to perform such tasks or assist workers can greatly reduce risk of human accidents. For example, in the wake of the Fukushima nuclear reactor meltdown, plant operator TEPCO deployed remotely controlled robots to investigate the highly radioactive environment.
Quality. Robots are excellent at performing repetitive tasks with defined accuracy, that is near-perfect accuracy: in assembly lines for mass production, machines can significantly increase yield by consistently manufacturing products that do not need to be discarded or reworked.
Operational performance. Automation can bring significant cost savings when leveraged to reduce the number of full-time employees needed to operate machinery. Human workers can take on supervisory or more cognitive work, while robots can function continuously without needing shifts or pauses.
Flexibility. Compared to traditional manufacturing machines, robots offer increased flexibility of use with mobility, modularity, and environmental awareness. They can be reconfigured by software and component changes to adjust to new production requirements, without requiring substantial capital investments and factory floor redesign.
Where it is today
The use cases of robotics can be roughly divided into two categories: industrial and professional services.
The former consists of both manufacturing machines performing repetitive tasks along an assembly line, and robots co-located and collaborating with workers without additional safety equipment. The World Robotics report by the International Federation of Robotics shows that in 2018, 64% of the global robotics hardware market was dedicated to industrial use cases. From 2014 to 2019, installations of industrial robots increased by 11% CAGR, for a total of 2.8 million units in operation. Japan is the world's second largest industrial robotics market and, despite its advanced level of automation in industrial production, has been sustaining a strong growth rate of installations.
The automotive industry is a strong adopter of robotics, with almost 28% of total installations. Universal Robots, a young leader in industrial robotics, has collaborated with several car manufacturers, from the PSA Group to BMW, to deploy cobots on the assembly line. PSA leveraged collaborative robot arms to build 200,000 cars in the first year, with zero failures, significant cost savings, and elimination of personnel exposure to chemical fumes.
Electronics is the second largest robotics customer industry thanks to the sustained demand for electronic components. Apple developed the “Daisy” robot to automate the disassembly of iPhones into individual components and recover materials: at a rate of 200 phones per hour, each Daisy can disassemble 1.2 million phones per year, helping divert 48,000 metric tons of electronic waste from landfills in 2018.
The second category, professional services, comprises a variety of commercial applications, from support in warehouses to healthcare and retail. In 2018, service robots captured 36% of the robotics hardware market, and that share is poised to grow to almost half by 2025. Still relatively young, the market for service robotics is highly fragmented across segments, with large incumbents coexisting with startups and digital players. The vast majority of players is in the US, but Japan has the fourth largest number of service robot manufacturers globally.
Logistics robots help make assets available quickly and efficiently, in warehouses, hotels, and offices. In 2019, they accounted for the largest share of sales (43%), supported by the boom in e-commerce sales triggered by the COVID-19 pandemic. As factory or warehouse AGVs such as Aethon's TUG become more popular, interesting solutions for other workplaces are also emerging, such as the Savioke Relay, which can be customized to safely deliver documents, supplies, and even medical samples.
Patient outcomes-driven innovation has resulted in the commercialization of more robot-assisted medical technology. Robotic arm assisted technology from Stryker, for example, combines 3D CT-based surgical planning software with a robotic arm that operates precisely according to that plan, preventing unnecessary tissue damage, obviating invasive surgery, and leading to better patient outcomes in hip and knee surgeries.
Globally and in Japan, it is expected that the adoption of robotic automation will continue to increase and improve upon current use cases, as well as find new applications in new sectors. The four main drivers for growth are as follows:
Improving economics. As labor costs increase and robot prices fall in most economies, automation becomes more desirable: a factory can now break even on the capital expense for a robot, compared to the human cost for a given task, in under a year in many situations. Moreover, continuous efficiency improvements, with further associated cost savings, will keep creating incentives to replace machines more frequently than before.
Labor shortage. When economies develop, the range of better-paid, safer job alternatives widens, and the number of people willing to perform repetitive physical labor decreases. Along with Japan's ageing population, the demand for robots to provide such forms of labor is bound to grow.
Increasing automation scope. As technical advancements lead to robots with broader abilities, greater safety and autonomy, the range of tasks that can be cost effectively automated will grow.
High flexibility. Expanded diversity in production calls for more flexible setups, in order to prevent expensive retooling and optimize the utilization of the factory floor. Multiskilled robots that can be quickly programmed for multiple lines help reduce such costs and speed up readiness for production.
How the technology will continue to evolve
While hardware advances will create lighter, smaller, and more durable robots, software innovations have the potential to deliver a truly generational leap to robotics. Focusing on this domain, there are three main areas of development.
Increased autonomy. From simple rule-based control and planning software, robotics has made a generational leap due to AI. Thanks to reinforcement learning and other cutting-edge ML techniques, computer vision and reasoning will vastly improve robots' ability to recognize objects, avoid collisions, and move more smoothly around diverse spaces. Building on Industrial IoT and cloud analytics, it will be possible to deploy robotic fleets that communicate to each other to maximize efficiency and report their status to the cloud to enable more granular predictive maintenance.
Programmability. So far, instructing robots has required significant software capabilities, and reprogramming them is often a time-consuming and machine-specific effort. "Low-code" and "no-code" platforms promise to create intuitive robot command interfaces, which will allow operators to tweak a machine's operation in a shorter time. Startups such as ArtiMinds and drag&bot are offering such systems, and more solutions are expected to emerge.
Human-machine interaction. As collaborative robots gain popularity on and off the factory floor, human-machine interaction and "social robotics" technology will improve and find commercial application. In order to facilitate interaction with untrained personnel, customers and passersby, it will be necessary to create robots capable of understanding and producing human language in dynamic contexts. "Affective robots" capable of responding to human emotions, and "grounded language learning", the task of inferring the meaning of words from the surrounding environment, are still topics for academic research, but will surely make progress in the next decade.
The key future applications
It is possible to assess various sectors' potential for automation by considering the level of expertise required, the extent of predictable and unpredictable physical work, the data available, and the necessity for execution. Accommodation and food services show the greatest potential, driven by the large amount of predictable work involved in hotel management and food preparation and establishment maintenance; next, manufacturing and logistics, driven by the automatable physical work and large amounts of data; fourth is agriculture, which will require more dexterous robots to deal with the larger amount of unpredictable work.
Looking at industrial robotics in particular, one of the most challenging yet value-unlocking developments is "jigless welding": while traditional welding robots require a human to mount and unmount components onto a jig in order for them to be welded correctly, new welding robots are able to autonomously grab the components with one arm and weld with the other. Robots such as those developed by Yaskawa can weld "on the fly", thus vastly increasing throughput time.
The synergy of robotics with other digital disciplines presents the potential for Japanese players to further improve the impact of automation in industrial settings. For example, ML can be leveraged to improve the operational accuracy and effectiveness of predictive maintenance in applications such as welding, painting, and assembly. Cloud infrastructure can be used to create centralized information management systems, through which robots can report and resolve issues and help machines plan and coordinate with each other.
Still a relatively unsaturated market, service robotics in Japan is expected to grow significantly at a CAGR of 14% between now and 2035, driven by such technological advances and by the labor shortages created by an aging population.
To support the elderly population, there is likely to be an increase in use cases for mobility robots, such as WHILL's robotic wheelchairs, and healthcare-related solutions for hospitals and nursing facilities.
Facility management applications show promise as well. At the moment, the most common use case is cleaning robots popularized by iRobot, but there are interesting opportunities in security robots, such as ALSOK's REBORG-Z, which can patrol facilities, provide support, and detect anomalies. In 2020, Toyota demonstrated a ceiling-mounted robotic arm, which can support in house chores such as loading dishwashers and cleaning.
Logistics robotics is a market of growing interest, thanks to the sector's high automation potential, along with the growth of e-commerce. However, many of the high-potential tasks in logistics also require sophisticated technology, such as picking and placing, as well as return handling, which is currently labor-intensive. New players, such as Magazino in Germany, have started developing dedicated solutions, and Japan, too, is in a prime position to compete: having spawned several players with advanced AGVs and robotic arms that, paired with advanced machine vision solutions, could build powerful products to compete in this space.
Given their consumer-facing nature and broader, untrained user base, service robots face a much higher bar in terms of expected functionality, ease of use, and safety. Creating robots that people will appreciate on a daily basis will require an intimate understanding of users, as well as the ability to iterate quickly without excessive go-to-market times and costs. In order to maximize their success, players in Japan should embrace design thinking techniques, as well as adopt an agile mindset towards robot design and controller software development.