Mobile Robotics and Robot Toys

Background

Mobile robots are designed to work at a distance and, if autonomous, to solve problems without outside assistance. The implication of this is that if they are to perform important tasks, we must minimise the possibility of them being made useless through decalibration brought on by collision with objects and other hard knocks.

JIGSAW and C-SAW were originally designed for mobile robotic applications. Together, they allow a robot to integrate its array of sensors, then map its motor commands to the composite 'image' formed by its sensors. This is done without any prior calibration, and can run continuously in the background to  ensure that even when sensors get misaligned, or motor systems are distorted, the robot continues to function competently.

Opportunity

Within general mobile robotics, several toy opportunities remain unexploited:

A new kind of toy robot: assembled by children, but able to really learn and grow smarter naturally.
Robot construction kits that once assembled are able to perform smart object recognition and fine autonomous motor control steered by sensory feedback.
Doll-like robots that are shipped in uncalibrated states and spend the first part of their lives going through real learning and coordination development.

Current Technology

However it is done, robot calibration compromises robot autonomy and adds costs. Current options:

Use of calibration bureaux
Regular calibration phases requiring supervision and special environments
Highly constrained high-cost manufacturing processes with all their limitations

JIGSAW benefits – short term

JIGSAW has the ability to make cheap robots smarter and smart robots cheaper. As sales volumes grow and unit prices fall, costs and dependency on servicing in all fields of robotics must decrease. JIGSAW has the ability to derive calibration of mechanical and sensing equipment even when the design and specifications of the equipment are unknown (and changing).

Once a robot has been assembled, in the factory or on the playroom floor, JIGSAW lets it discover for itself how its various parts (cameras, joints, etc.) work together. There are normally plenty of costs attached to this process, but JIGSAW makes it cost-free.

JIGSAW simultaneously increases durability, makes behaviour more life-like, and reduces costs. These benefits come as an integral package, not a trade-off.

JIGSAW even supports manufacturing processes where the final design is unknown. JIGSAW is flexible enough to support a production line on which every product is as individual in configuration as human beings are in shape. The analogy is apt, because JIGSAW copies important aspects of the developmental processes of our brain.

JIGSAW is an automatic process that improves the way robot perception and action are integrated. Unlike standard interventionist methods, JIGSAW lowers costs, increases life-long reliability and enhances robot behaviour.