Junji Suzuki Manager of Laser & Electronics Group, Turbomachinery & General Machinery Department, Machinery Headquarters

In September 2005, Mitsubishi Heavy Industries began selling the world's first household robot. Called wakamaru, the robot communicates with persons and can distinguish between ten different faces. For generations, just a fantasy of science fiction, today, the home robot is a reality.

What led you initially to develop a robot?
In June 2000, our president asked everyone in the company to submit ideas for new lines of business and products. Over 1,800 new ideas were submitted.

  Team members came from company offices around the country to pick up the real gems from the mass of suggestions. Among the responses was an idea for a home robot. So, from the start, the project was not about producing a mascot or a publicity stunt, but rather the pursuit of a profit. No one in our team was a robotics specialist. Our first task was to go around the company collecting whatever existing technology we could find that was applicable—and what we found was image-recognition and robot arm technologies from our manufacturing divisions. Then, with the cooperation of the few robotics experts in the company, we started development at our Kobe Shipyard & Machinery Works.

What was the hardest challenge you faced in developing the robot?
We envisioned a robot that would provide many kinds of help and be integrated into the Japanese household, which today is increasingly nuclear, and frequently with two working parents. The biggest question then became what would give the owner the greatest satisfaction? This is what we spent the most time deciding. We also discussed how the robot should speak and act, which is to say, what type of personality would be best. But, it was the first time for everything, a complete process of trial and error. We reflected; we studied; we talked with and listened to other companies and universities. Each time we hit a wall, we tried a new approach—over and over.
  Since we are not an electronics company and knew we could not fulfill all the technical requirements with in-house capabilities alone, we sought outside support from the start, and I think this contributed to our success. The company's high expectations created pressure, but we were able to focus completely on the project, and the young engineers enjoyed themselves. These were also factors in our success.

What is the robot's most noteworthy feature?
No particular feature is that exceptional. But the robot's ability to exist in the home without being at all bothersome is a great merit. Left alone for 24 hours, wakamaru will take care of its own needs, stroll between rooms and recharge itself as needed.
  Our next goal is to improve its ability to pick up and hold objects. But identifying objects and then holding them properly are highly challenging functions.
  What we do not intend to do is make the robot more humanlike. The robot exists to help humans by interacting, but it is only a robot, and will continue to evolve as such.