The applied robotics industry has experienced a number of waves of change since its inception in the early 1970s, and 2010 represents the groundswell of the next major change.
The first successful industrial robot was the play-back robot: a program flexible manipulator that played-back precisely and untiringly the sequence of moves it was taught. It started with the Unimates of the mid 1970s; these were hydraulically-actuated because electric motors at the time did not have the power density to be useful for medium-to-large robots. Over the next 20 years, electric servomotors took over as the robots' primary drives, offering greater serviceability and eventually more power and control.
Each generation of development not only gave improvements in performance, reliability and control but, due to the maturity of the contributing technologies and the increasing robot production volumes, also provided massive gains in performance/cost ratio.
Nevertheless, the bulk of robots now in production use are play-back robots, which in the industry parlance work within a 'highly structured' environment. Everything in this environment - the workpiece and the workstation - are precisely known, predictable and do not change.
For example, every production manager is aware of his many sophisticated and high-speed processing machines which, though being automated in themselves, still require manual loading at the infeed end. This is because the most common and economic means of bulk presentation of the inputs to these process machines is an 'unstructured input' to the current generation of robots; in other words, the robots that cannot cope with leaning, twisted or flowering stacks, or approximately-placed workpieces, on an infeed pallet.
The next wave
Extending the robots' capability to embrace these less-structured areas of manufacturing environments is the next major wave of change.
Already, the beginnings of this next S-Curve have been happening since the mid 1990s when pioneers in the technology started to add sensory technologies that enabled the robot to make small adaptations in their pre-programmed moves to account for variations in this structured environment.
Examples of this adaptive capability are: the seam tracking welding robot, the use of early vision systems to 'see' the precise orientation of a workpiece so that the robot can adjust to it; and the use of sensors to perform rudimentary quality control so robots can sort the articles they are handling.
As the sensory technologies increase in capability and speed of processing, in real terms become less costly, and become more and more integrated within the robot's controls, this combination will fuel the rapid adoption of this next genre of adaptive robotics technology.
First among these new sensory technologies will be increasingly able and faster Vision Systems; leading the charge will be the integrated vision and distance sensing systems already being marketed by leading sensor companies.
Omni-sensors that will image the entire workspace, coupled with adaptive and predictive computer modelling afforded by increasingly powerful, fast and low cost computers, will give these robots a more precise awareness of their variable work environment than was ever available for human operators.
We are not talking about robots working in totally unstructured environments, as even when the technology eventually becomes available, it will never be efficient enough to run a manufacturing line with that degree of disorder.
Not only will these robots cope well with the simple task of de-stacking an untidy pallet, but their new capability will open up new swags of manufacturing tasks in which the workpieces themselves are "unstructured" in that they change their form continually, or simply vary greatly from one to another. These are the pliable, soft, elastic and limp workpieces such as fabrics and soft polymer products, or the low-tolerance assemblies such as wooden crates.
A way to go
Of course, for robots to perform effective work in these new areas, there will need to be accompanying developments in manipulators such as multi-armed robots, as well as in their end-effectors such as super-dextrous grippers, sensory grippers, grippers that can handle limp and porous materials and fragile workpieces. The ability to perform real quality control as an embedded capability will also be important, particularly if this function is integrated with the robot's manipulation functions.
Clearly, imbuing robots with such peripherals in a cost-effective manner will enable their scope of usage to be greatly expanded.
The technologies that are needed are more evolutionary than revolutionary: indeed, much of their development is already well under way. Like the robots themselves, these peripheral technologies will advance to a point, perhaps in the next five to 10 years, when they will be fast and capable enough. Then, with increasing usage starting in high value applications, their effective cost will reduce so that once the critical mass point is reached, their affordability will make them commonplace.
At the moment, the performance/cost ratio is holding back many potential applications - only the high-end applications are being implemented.
Dr Paul Wong is managing director of Applied Robotics.
Add a comment