The year is 2018 and there are some pretty impressive robots out there. Robots can now run, jump, fly and think in ways that were sci-fi just a few years ago. See Boston Dynamic’s latest video if you are any doubt of the technology now available. Collaborative robots like the ABB Yumi, Universal Robors UR series and Aubo i5 are transforming the way we work in manufacturing. Big data analysis and artificially intelligent processing are changing the way we think and strategise. So what’s next? Here are a few ideas of what is likely to happen in the world of industrial robotics.
Motors: The last few years have seen motors become smaller, more powerful, more economical and more controllable. This trend towards power density will continue allowing faster lighter more agile robots and the price will continue to drop as volumes increase. Add to this the huge improvements in motor control and you have the ability to do much more, much better, for less. However there is also new technology that may yet usurp the status quo of motors – the combination of motors with magnetic gearboxes has been around for a few years but developers have been struggling with making them strong enough to deal with the massive internal forces of prototypes, let alone make them mass produceable at volume. This is analogous with the start of the jet age: A fundamentally new technology needing materials and design technology to catch up to turn the theory into reality. With the possibility of creating new machines a step change better than their predecessors – not convinced? Take a look at Genesis Robotics.
Gearboxes: Gearboxes are the unsung heroes of industrial robotics. You can make something move like a robot sure, but repeatability and accuracy? That’s down to gearboxes with very low backlash or slop. The cycloidal (such as Sumitomo) and strain wave (Harmonic Drive) gearboxes have been at the heart of robotic development for the last few years. They deliver the high gear ratios, huge torques and accuracy in compact forms that make today’s lightweight, slender robot arm designs possible. One issue over the last few years has been the cost of these precision units. They are not easy to make or indeed copy. However many new designs are on the brink of reaching the market, new twists with new patents on the high gear ratio, high accuracy torque multipliers. This and the increasing production volumes of industrial robots are pushing prices down. Expect gearbox costs to halve in the next few years with a significant knock on to robot prices.
Motor Control: Gone are the days of massive heatsinks needed to absorb unwanted power from motor control systems. This has allowed motor controllers or ‘drives’ to become much smaller, even mounted within the motor housing itself – simplifying cabling and reducing the need for bulky separate control boxes (see the Meca 500 with the controller built into its diminutive frame). Motor control is becoming faster and smarter – intelligently varying power depending on many parameters. This is the technology behind drones being able to fly for longer and while industrial robots are not often reliant on batteries (with the exception of AGV’s like Amazon’s) the ability to use less power, run quieter, and more powerfully is very important in pushing development. More than this intelligent motor controls allow for systems to take out wear and backlash and allow the sensitivity to unexpected loadings that make collaborative robots possible.
Sensing: Many sensing technologies have been available for a while but have had somewhat restricted use. This has been down to two factors, cost and the ability to use the information. Intelligent filtering of data along with lower cost sensors means that tomorrow’s robots will not be dependant on this switch or that sensor to work out what is happening. They will parse many data streams and be able intelligently process them into useful actions. A robot system will not be reliant on one level of safety but have multiple streams that can verify each other making robots far safer and easier to implement. Sensors that deliver a lot of data such as Cognex vision systems are becoming even more useful with the ability to process and use that information better. Dumb robots will disappear as this technology, also being developed for autonomous vehicles, is adapted wholesale.
Programming: Programming of industrial robots remains a bottle neck in their implementation. Industrial robots are very capable multi purpose machines and as a result very complex. Robot programming languages can contain thousand of specific commands. Whereas many computer languages have been built into reusable blocks that can be then used to create more complex programs industrial robots are still frequently both dumb and complex – very powerful yet difficult to give simple instructions too. Lots of room for improvement then yes. And things will start to happen quite quickly in the next few years as the way we interact and instruct robots is set to change. It has already started but the teach pendant still has a grip in the industry as more intelligent and comprehensive computer modelling will allow complex programs to be made at least semi automatically by software. Including, vitally and at long last, better incorporation into the CAD CAM (computer aided design computer aided manufacturing) ecosystem. Add to this the increased safety and a better general “machine intelligence” it is possible to envision a robot system that gets all the information it ever needs from the design stage to make clever, safe, adaptable and flexible production choices – more like human workers.
Our interaction: So will these all conquering future robots, the Fourth Industrial Revolution, drive us all into unemployment? Certainly this can and does happen on a local level and there is not much point denying it. We can expect to see an increase ‘churn’ of employment as industries are disrupted on a wider scale than the introduction of cotton mills or the motor car had in the past. However on a macro scale the data tells a very different story. Companies, industries and countries with high robot use are measurably the most productive and successful. This is not just in terms of profits for factory owners but also in the wages, lifestyle and indeed sometimes numbers of those employed. As the simple unwavering economic logic of using the best, most efficient form of production slowly crystallises in our minds hopefully we will start to see robots as more than a threat to jobs and more as the right tool for the job: To help free us from workplace drudgery and servitude and move us into a brighter future where we can have more aspirational and creative callings for ourselves and future generations quite possibly in industries that do not yet exist. Still not convinced, well, be en large, the means of production have been concentrated in the hands of the few for the last few hundred years. The next generation of robots will be more affordable and useable for more people than ever before and that is hopefully a good thing.
Sumitomo Cyclo Series
Mecademic Meca 500
World Economic Forum
International Federation of Robotics
Nesta Innovation Foundation