A helping hand
Advances in vision and safety systems and the need to cut costs are driving manufacturers to implement the latest robotics on the factory floor.
As we emerge from the brief hiatus of Christmas and brace ourselves for the economic storms forecast for the year ahead, this is perhaps not the best time to be talking about investment in manufacturing.
Orders are drying up, credit's evaporated, exchange rates are in meltdown and - in the UK at least - cheap labour is heading back home, as their domestic economies approach a par with that of their temporary home.
Conventional wisdom has it that it would be ill-timed, reckless even, to be spending any few spare funds on new or upgraded equipment, in this case factory-floor robotics and automation. Yet this is exactly what appears to be happening, at least according to robotics suppliers.
And not before time too, it seems. According to figures from the British Robotics and Automation Association (BARA), since 2000 the annual number of industrial robots installed in British industry has fallen from a high of nearly 2,000 to fewer than 1,000. As BARA chairman Professor Ken Young explains: "Britain has not invested in manufacturing development over the past eight years or more, and has spent all its money on cheap labour instead of automation."
Driven from behind by the need to cut costs, some manufacturers are also being drawn towards increasing automation by recent developments in the technology that offer greater flexibility and cost-effectiveness, with advances in robotic vision and safety - and integration of the two - allowing greater collaboration between the robots and their human operators.
Engineering manager at Fanuc Robotics UK, Mark Arbon, says: "Ironically, the credit crunch has provided industry with a spur to automate. Our level of business remains steady and there's no sign of a drop-off in the near future. Loss of cheap labour is a factor but we've also seen investment in arduous applications such as handling concrete slabs, so health and safety has been a main driver as well."
Another factor is initial outlay. For example, the cost of a six-axis industrial robot is reported to have fallen by 50 per cent in the past 15 years.
And Arbon agrees with Mats Myhr, product manager of robot control systems at ABB Robotics, as well as systems engineer at Pilz Automation Technology Stewart Robinson, that these new-generation systems are being adopted in all industrial sectors and by all sizes of company.
Innovation in the software
On the vision side, the most recent advance has been the advent of 3D vision systems, says Vishnu Sivadevan, industrial automation technology analyst at consultancy Frost & Sullivan. "These have eliminated many of the traditional obstacles to the adoption of vision guidance systems for industrial robots," he says.
"For example, the presence of shadows or highly reflective surfaces in an image could not be handled by some earlier vision systems, so higher levels of illumination were needed. But this is no longer the case thanks to improvements in image processing software and the advent of 3D vision sensors and cameras."
He says that, in general, these sensor-based systems are more attractive than other approaches such as laser triangulation because they are simpler, allowing the cameras to be more compact, and cheaper because the sensors are off-the-shelf components.
However, he points out: "Sometimes the real innovation in a 3D vision system might lie entirely in the software, which would enable the system to give greater performance compared with its peers. While 3D cameras provide image acquisition of good quality and detail, image processing algorithms play an important role in feature extraction and recognition operations from the images."
On the safety side there is general agreement that the key underlying advance has been the replacement of electromechanical relays for control logic with safety PLCs operating over safety-related control networks such as SafetyBUS p.
By doing away with the need for hardwiring of relays, these new controllers and networks have made modern industrial robotics systems more flexible, more modular, and easier to test and maintain - and cheaper too because design, installation and commissioning costs are lower.
Sensor technology has come on apace as well. Light curtains and laser scanners, for example, can detect human presence in a dangerous zone, while an emerging technology called sensor fusion - merging data types from, say, accelerometers and gyroscopes - gives greater control over a robot and therefore faster reaction times.
The ability to integrate these devices into a safety-related network means functions such as monitoring standstill, reduced speed and position allow safety-related uses that may not have been considered earlier or may have been prohibitive for technical reasons. "For example, it's now possible to define safe areas and/or safe robot speeds [in the system software] depending on inputs from safety devices, prohibiting robots from entering particular areas or operating at high speeds when these areas are being accessed by human operators," says Fanuc's Arbon.
Integration is the attraction
But it's the integration of vision, safety and control that is proving the real attraction of these systems, and many of the main robotics suppliers have their own offerings - for example, ABB has SafeMove, Fanuc has Dual Check Safety (DCS) and Pilz has SafetyEye, all of which achieve the same end but with differing approaches.
Broadly, these integrated systems do away with many of the current safety devices such as light curtains and protective barriers. For example, as Myhr explains: "Vision systems are now safety rated and can thus be used as components in the overall safety system of a robot cell, which means they can supervise huge areas and
partition these areas into numerous zones.
"Combining safe vision with safe robot control gives the ability to reduce a robot's speed or go all the way to a temporary standstill, while it remains in automatic mode, if someone comes too close to danger. As soon as the person is gone, operation can quickly resume at full speed."
This new degree of control has two key effects on the factory floor - more compact robot cells and less downtime.
Now that safe areas and/or robot speeds can be defined precisely, safety distances can be reduced, allowing more efficient use of floorspace. And because these spaces - called safety-rated soft limits - are software-defined they can easily be changed according to fresh needs.
Also, since functional safety and motion can now be combined, a robot does not usually have be stopped completely for maintenance, for example, which is costly - of the order of thousands of pounds a minute in sectors like the automotive industry - if, as sometimes happens, the robot has to be reset.
All of these factors promote closer human-robot collaboration or interaction, and are allowing new ways for the two to work together. For example, as Arbon says: "Operators can now safely load parts directly to fixtures being carried by the robot without the risk of the robot moving unexpectedly. Also, routine maintenance can be carried out on end-effectors such as welding torches, spot-weld guns and so on, again without the risk of the robot moving."
Goodbye to barriers
Yet these new-generation systems are not suitable for all applications, nor will they completely remove the need for barriers. Robinson at Pilz says: "Suitability is a key issue, and these systems would not be used in processes where there's a danger of parts being ejected. And for areas where the danger area is moving around, as with gantry-mounted robots for example, issues remain."
Myhr adds: "We must acknowledge that these new safety systems do not make sense in every robot automation project, such as spot welding in an automotive line or where repeat movements are needed, as in palletising."
As for saying goodbye to barriers, Myhr says: "I think barriers will still be there to some degree in most cases. To have complete openness in all directions does not necessarily make sense, since you still have to respect the proper safety distances, which means you will occupy a lot of floor space - possibly without obvious benefit. I know of only one totally open cell at the moment, a lab set-up which is due to go into real production."
But truly collaborative robots are on the horizon. One sign comes from research by Japan's National Institute of Industrial Safety, which in 2005 proposed a concept for the safety control of a cooperative robot carrying out materials handling duties inside buildings under construction.
Dr Richard Piggin, chairman of the Safety Special Interest Group at BARA, tracked this research and says: "The NIIS safety system used layer of protection analysis (LOPA), a risk assessment tool used in the chemicals sector, which provided a stratified number of independent safeguards called independent protection layers and operating in a time sequence.
"Also, a principal feature of the control system was a programmable safety controller for safety-related motion control that was independent from the general-purpose PLC. This was just a proposal, though, and at the time further research was recommended."
Not so far off, however, are further developments in the current technology. Arbon cites wireless safety devices as one example: "Not readily available yet but likely to be developed," he says. And Robinson says: "I think that well within five years from now we will see gantry robots with these safety systems. Also, we'll see systems with greater capabilities for detection and differentiation between objects, greater range, faster system response times and more detection zones - all of which will be advances in the base technologies."
Frost & Sullivan's Sivadevan agrees with this point about base technologies, adding: "Broadly, we're seeing the start of the era of 'service' robotics, which will usher in the age of stationary robots working with humans without fences, and mobile robots that safely run errands to move goods in a warehouse or on a factory floor."
These may be grim times for the world's economies, when securing funds for any new technology will at best be difficult, but companies who have invested in these new, flexible and cost-effective approaches to automation could well be the ones who are best poised to survive the downturn - and profit most from the recovery when it finally arrives.
Now all we need is for someone to talk some sense into the banks.