Tatsuya Matsui & Hiroaki Kitano, with ties to SGI (Silicon Graphics Inc) both believe that a robot’s exterior appearance will dramatically affect how humans perceive and interact with them, including whether or not we will accept them at all.
SIG was never completed, but was designed with the rule of thirds in mind, an ancient artistic measurement which defines beauty. During SIG’s development, it was determined that the female body is roughly the right size for a robot to coexist with human beings. If a robot is too small, it will appear child-like which may betray its function. If a robot is too large, it will not be adopted by the public for fear of it overpowering the space it occupies. While SIG is destined to never leave the laboratory, she was instrumental in shaping Matsui-san’s philosophy. Since leaving SGI to form his own company, he has stressed the importance of outward beauty complementing the hidden inner complexity of his robots. Had SIG been completed, it would have stood 168cm (5’5″) tall.
The following is an excerpt from Tatsuya Matsui’s website (no longer online) about his work with JST ERATO and Hiroaki Kitano’s Kitano Symbiotic Inc.. It has been reproduced here for educational purposes for those interested in the history of his work.
SIG the humanoid: Prototype of Robot Design
What person is the robot?
What is it exactly, or rather, who is it we are designing ultimately comes into question. The first point we have to consider is that the robot tries to think from the point of view of an instrument. From that point the approach broadens the paradigm of conventional industrial design. Industrial design, we must remember is the design of an instrument. It is the design of objects whether they are household appliances, cars, space shuttles or even the most basic tools like the hammer and saw. However, even these instruments are going through an evolution of their own as we began to produce and use these instruments as tools necessary for our very survival. Once animal bones and stones sufficed for this purpose, however it was the invention in the middle of the 20th century of the digital computer that was fundamentally different from conventional instruments.
The computer was the first instrument to operate information in the place of the object; whose function was not physics but rather the abstract concept of programming. Although this has afforded a certain degree of freedom for the designer it has also meant simultaneous chaos, resulting from the separation of the one that performs the function from the idea which has occurred. Design must be approached from the view point of how it can serve as a hand in enabling clear, easy access to the user and put into context, the assemblage of chaos within. In his book “the Psychology of Everyday Things” D.A.Norman examines the role of interface design in communicating the function of the object (and it’s desirability as a product). As much as design had to assume and be equal to the complications that arose with the appearance of the computer, so it must again broaden its role in the genesis of robot engineering and technology.
The autonomous nature of the robot creates a division between itself and more conventional instruments. Interaction between human and present robots is still far from satisfactory, however the acceleration of research in this field makes it a matter of time before the interactive possibilities of the robot will clearly separate it from mere instruments. What is inside the robot will be largely influenced by its appearance. What kind of relation it will have to its structure is a theme that has to be explored. This is a problem particularly applicable to the non-humanoid robot.
An example of this kind can be shown in the development of AIBO. Those who participated in the SONY CORP project from the beginning, first visualized a kind of insect, as six legs facilitate the smoothness of movement. But since the first prototypes conveyed less an insect than the universally reviled cockroach, a substantial design change was necessary. While the four-legged creature lacked smoothness in movement, it was not however, a liability. In fact, the awkward, rambling gait of the earliest prototype was decisive in SONY’s decision to support the project. Robot development in this field, it was decided, cannot be viewed as the development of alternate animals but rather as robots having similar attributes but functions entirely separate from those of living pets. Although AIBO is often called a pet robot it is more accurately in the realm of digital creatures as reflected in its design which doesn’t shy away from its mechanical functions. On the contrary, digital creatures are by design a sum of their mechanical parts.
The same is true of the time the humanoid robot was introduced into the Kitano project. During research into human intelligence the first humanoid would have had the appearance of a human through the applications of silicone and cosmetically enhanced features. The awareness however, that project members were pioneering a whole new field of industrial design was instrumental in rejecting non-essential visual aids on the prototype. This would have gone against the spirit of the project, which was again not replicating a human being but analyzing the structure of a human’s cognitive functions.
It is the main goal of this project to explore the basic theory of human intelligence. Therefore the visual and auditory capabilities of the autonomous robot must be equipped with visual and auditory features aligned much the same way a human’s are in order to assess their integration into our cognitive functions. The robot must have the dimensions equivalent to the upper half of a human body to avoid diffused reflection from visual and auditory sources that can result when the structure deviates from that basic form. It was discovered that the processing of external voices became very difficult, making it necessary to attach a cover over the auditory mechanism of the robot to block the sound from outside. This combined with the sound of the motor which occurs inside the robot- another interference, was decisive in introducing an element of design. It is to understand the intelligence of the human being that makes the humanoid a necessary element in our research, however, the nature of the project cannot be expressed sufficiently by the mere replication of a human.
The structure of the robot being similarly adapted to the human from is one option towards efficiency. Design plays an important role in opening up a whole new field of inquiry; one which all researchers can value as a visually expressive validation of their work. This particular robot will remain forever consigned to the laboratory for experimental purposes. Here the robot’s dimensions will signify the entertainment humanoid robot’s place in the consumer’s home. At present, the research robot sits in a 12 square meter room crowded in by two desks and at any given time two or three researchers. The entertainment robot will therefore have to adapt in much the same way to its surroundings. Its size will be an important consideration when consumers decide whether or not to allow it into their homes, and will most likely reject it if its dimensions threaten to overpower the space it is meant to share. The robot must settle as compactly as possible. This is particularly relevant to Japanese houses, with an average living room size of 12 square meters. The impression of distance is often measured subconsciously.
A robot ‘s proportions however, must adapt to its height. Therefore, a compact, child sized robot uncomfortably conveys its function as a toy so whereupon a small system is desirable it must not obfuscate its meaning by turning it into something entirely unintended. We have judged that an adult woman has the dimensions desirable for this purpose and can accommodate the small systems designs that already exists for this purpose. Although small in stature, it suggests the form of a human being ideal for use in limited space. The research robot used in the this project has been designed with arm sockets despite not having any mobility in the body to entice the imagination into forming a complete picture; a playful homage to the Venus de Milo whose very lack is at the center of her embodiment of complete beauty. This is how design is succeeding as an interactive medium between the viewer and the object. The imagined product at eight heads high has its roots in classical Greek sculptures. Ideals of feminine grace, dating back as early as the 5th century B.C. have its roots in the Golden module which determines the 3:5:8 configuration as the fundamental basis by which the body is measured and proportioned. If the face is measured vertically at 21 centimeters and again from the chin to the top of the breast down to the navel, these dimensions multiplied by eight comprise the ideal by which beauty is measured and by which we can determine the robot’s dimensions as an ideal. This is how we formulated a concept by which our aesthetic standard evolved.
I drafted a sketch for a two-dimensional research robot which was sent to an industrial mold maker and within days the three dimensional product was realized. When the cover was at last installed over the mechanism, the effect was unsettling to say the least when the robot ventured its first movement. Here for the first time intent was expressed through the movement of an object. This is where the full realization set in that this is not the enterprise of object design we were engaged in but more to the point the design of a whole new range of dynamics. The robot developed by the Kitano project, while purely for research purposes, marks the beginning of what will one day be known as an ‘entertainment robot’ that SONY through the development of AIBO has already begun marketing. Already, there are indications that the possibilities arising from such a venture will further increase as more enterprises get involved in this business in the 21st century.
The humanoid robot will take on a significant role in the aid and care of human beings as well as performing more dangerous tasks. From this point on consumers will largely determine and influence the standards that will be set as to the merits or demerits of its design if the production and design of robots is to become significant as an industry in the future. The research of robot design will be carried out simultaneously with the technical research involved in such a project. It is necessary therefore to enter into the production of robots by establishing a ‘design laboratory’ to integrate the data and research results into the design and production of entertainment robots. In much the same way that cars and computers became the industrial base of the 20th century, we envision the entertainment robot as advancing a whole new technology. The role of design will be integral to this endeavor.