Until now, robots have been playing key roles in industrial settings such as factories and warehouses. In recent years there has been growing interest in the use of robots in the home, such as smart speakers and household robotic devices, but it will still take some time before humanoid robots, like those seen in animation and movies, will become part of our everyday lives.

In September 2021, Honda, which has long been conducting research on robotics including ASIMO, unveiled some achievements in its research on avatar robots. Honda is aiming to put its avatar robots into practical use in the 2030s to provide value to society as soon as possible.

A leader in Honda R&D’s advanced robotics research shares his vision for the future that Honda is striving to realize through its robotics research.

Currently, robots are used mainly in spaces isolated from people at places like factories. However, in the 2030s, I believe robots will play active roles in people's daily lives.

For example, moderately speaking, we will probably be living in a world where delivery robots are running around town, and robots are replacing some of the remote tour guides for remote travel services* which have become somewhat popular in recent years.

Also, we will start seeing more and more avatar robots in our society, just like the one Honda recently announced, which I believe will change concept of “mobility” for people.

Avatar robots are robots that can act as a “second self” of the user. We are envisioning robots that the user can use as a second self while being in different locations.

There are two kinds of value our avatar robots can offer: Maximizing the value of time by enabling instantaneous travel that transcends time and place, and enabling the user to perform activities beyond their own capabilities.

First, being able to control a robot that is in a different location means that the user can jump over the distance and eliminate travel time. We describe avatar robots as “expanding the potential of mobility into the 4th dimension” as they defy the constraints of time and space.

One way to reduce travel time is to make efficient use of automobiles, public transportation and air travel. Avatar robots can realize the same value as instantaneous travel. In other words, use of avatar robots can maximize the value of time for the users.

The other value is in augmenting physical capabilities. Robots can access dangerous places where flesh-and-blood humans cannot go, and robots can do things beyond human physical capabilities.

For example, robots can carry extremely heavy objects, or they can do detailed work that is difficult for people with weakened eyesight. This is part of the appeal of avatar robots.

In this way, we believe that avatar robots enable people to transcend the constraints of time, place and ability, and enable us to realize our full potential.

The second use case we envision is in emergency lifesaving support.

For example, if a person collapses, it may be difficult for people around the person to properly use an AED (automated external defibrillator) as they may be panicking or lacking confidence. In such cases, if a person with expertise in medical care could travel instantaneously and be there to support the patient immediately, more lives can be saved, and everyone including the patient involved in the incident would feel greater peace of mind.

I think being able to have someone with expertise travel instantaneously to the site to help people in case of an emergency is a big advantage avatar robots can offer.

Space development has been commanding greater attention these days. I feel that the Moon and Mars are getting closer and closer to us.

For example, there is a plan for lunar surface exploration. By having avatar robots perform tasks on the lunar surface, we can reduce the risk astronauts would otherwise have to take in dangerous environments. In other words, the lunar surface exploration missions can be carried out from a safe location.

Also, by using the idle time of avatar robots on the Moon, the people on Earth will be able to enjoy fun experiences such as touching and throwing moon rocks in a zero-gravity environment.

Isn't it exciting to be able to experience the “real” world, rather than having virtual experiences while being in virtual space?

The metaverse exists essentially completely within the virtual world. We are more grounded in reality and place our value in offering people the ability to enjoy various activities and experiences anywhere in the world by having their avatar robots in real-world environment.

Honda’s robotics research began in 1986 as part of various projects alongside research into aircraft, automated driving, and other technologies. The underlying passion is that Honda wants to use its technology to help people, and we want to create robots which will be there in people’s daily lives.

Honda believes that people play the leading role. By using our robotics technology to support people in their daily lives, Honda wants to realize a society where everyone can fully demonstrate their abilities, open up their own possibilities and live fulfilling lives.

When we say that “our robots will be there for people,” it is not about physical distance. It is more about being psychologically supportive to people.

For example, when a conventional robot moves around a facility, it would put on flashing lights or buzzer sounds as if it is saying, "Coming through! Watch out!” So people have to pay attention and look out for the robot. That is more convenient for the robot and easier for the research engineers who are working on such a robot.

Ideally, however, people and robots should be able to coexist while communicating with each other. This is exactly what we mean when we say, “Our robots will be there for people.”

Besides the way they move, as shown in this sketch, every aspect of our robots including their appearance and size has been designed with the idea of coexistence in people's daily lives.

That’s true. Honda’s bipedal research began in 1986, resulting in P1, P2, P3 and leading up to the creation of ASIMO in 2000. I believe we have grown together with ASIMO for nearly 20 years since then.

My primary focus in ASIMO research had long been on legs/feet. It was daunting research, but I was really thrilled when ASIMO ran the way I wanted. I have been truly experiencing the fun of “monozukuri (the art of making things)” at Honda.

For the past 20 years, ASIMO has been in operation every day somewhere in the world, and had a lot of opportunities to actually interact with people. That is why we have been able to learn so much.

If ASIMO were to bump into a person and almost fall over, it could quickly make adjustments and keep walking. But what if ASIMO fell? What if there was a small child sitting there? When considering various cases, it was difficult to guarantee complete safety, and there were still challenges to overcome.

We had envisioned the ideal of robots that could move autonomously like people while being in the same space with people, but we think that would still take another 10 to 20 years, or even longer.

It is the nature of research engineers that we always have a desire to use our technology to help people. We want to use our technology in some way to benefit our society as quickly as possible.

So now, instead of getting hung up on developing ASIMO as a perfect package of all functions, we would like to provide value to society as soon as possible by focusing first on individual functions such as leg technology or hand technology. My image is that ASIMO’s DNA will be inherited by such respective research activities.

For the time being, we will continue to develop robots with specific capabilities, but there is a possibility that in the distant future, they will be combined into a single humanoid robot.

The avatar robot’s hand is being developed by engineers who formerly developed ASIMO using know-how amassed through their experience with ASIMO.

Honda studied human hands in the process of developing ASIMO. Many avatar robots from other companies have grippers, but Honda is developing multi-fingered robotic hands. A gripper would be totally different from a human hand, so to synchronize with the movements of a human hand and enable intuitive operation, we first made our robotic hand closer to a human hand in terms of hardware.

ASIMO can open the lid of a water bottle and pour water into a paper cup. This is a simple task for a human, but for a robot, the force of the hand holding the cup needs to be adjusted as the cup gets heavier when more and more water is poured in. ASIMO featured the technology to use sensors at its fingertips to sense the state of what it holds and make adjustments to achieve the optimal force.

This ASIMO technology is now being utilized in the avatar robot in the form of the “AI support.”

Our avatar robot is not like a puppet, which is merely moving by being controlled remotely by the operator. It uses AI to provide support in overcoming the difficulties of remote control. The ability to switch between remote control and AI-supported autonomy is the key feature of the Honda Avatar Robot.

I started research and development of robotics at Honda more than 20 years ago, and most of that time has been spent with ASIMO. Today, I am pursuing the research and development of avatar robots while applying ASIMO technologies and experiences, but my people-centric approach has not changed at all.

When I was researching ASIMO's legs, in order to make ASIMO climb stairs, I climbed stairs myself desperately trying to figure out the center of gravity of my body.

The development of the avatar robot is also people research, because, for example, we are trying to figure out what kind of information should be given to the remote operator to make them feel what they are grasping.