It was about a year ago that I last posted on ‘cobots’ (see this blog post “Myth Busting Cobots”) and I thought it was worth revisiting the area to see if much had changed.
In reality, not much has but the hype and push towards ‘cobots’ hasn’t gone away. So in summary, where is the market today:
- Cobots are still only predicted to be 20-30% of the overall global market in 2025 but locally in the UK, they’ve yet to get above 5%
- Cobots are the panacea that will solve all your manufacturing problems; if you believe the marketing people…
- Cobots are still governed by ISO10218 and TS15066 (both of which have just started their review cycles)
- Cobots are limited in payload and speed partly because of the above and partly because the safety cases for them are still harder to make than traditional robots
- The prices of cobots has come down a little, but in reality a similar sized (and faster) robot will still cost less today
So if the hype around these robots hasn’t diminished, are they really useful? I’ve never believed that they don’t have their place in the factories of the future. But I’ve always cautioned against believing that they are the only robot that you would ever need.
My view has always been that for a great many applications, a ‘cobot’ is simply not the right approach. Not because the robot isn’t capable; it may very well be capable of the task. But if that task carries any residual risk to the operator which requires guarding (or separation) from the operator, then why would you employ cobot when a traditional industrial robot would suffice?
I often cite the example of a company who bought a Kuka iiwa (which is a great robot, we have several at MTC) because they wanted their operators to collaborate with the robot in a press tending operation. The problem was, the press itself defined the safety case for the system; it was simply too dangerous to put an operator near the press but a robot could be put there. Having a cobot in the system looked great, but wasn’t necessary because it couldn’t collaborate with the operator because they were excluded from the work area!
There have been a lot of ‘collaborative workstations’ being created and shown at various exhibitions and conferences in the past 12 months (we’ve have our own version of this at MTC which we’re developing with clients). Some of what’s been exhibited looks great and in all possibility could really aid productivity. But a lot of what we’ve seen rings alarm bells because although ISO10218 and TS15066 allow for human robot collaboration in certain instances, the latter also defines where a dynamic or static force applied to the operator isn’t generally acceptable (e.g. the head and torso). So whilst a cell with a robot in front of the operator in theory looks collaborative and ’safe’; if there is any chance of a collision between the robot wrist/end effector and the operator, the standards indicate that this is NOT a safe situation.
But this isn’t just my view, at MTC we’re working with a number of companies who have installed ‘cobots’ but have not been able to make an adequate justification to their health and safety team, and so are not able to use them. We’ve heard of companies buying collaborative robots but having to guard them for even the most benign of tasks and of companies who have installed a ‘cobot’ but haven’t achieved the production rates they expected.
Fortunately, it’s not all doom and gloom, because there are examples in operation today that are successful and most importantly, truly collaborative.
Yet none of this is new; we’ve known about all this for a few years now.
In fact, the only thing that I believe is new over the last 12 months is a subtle shift in the robotics community from ‘collaboration’ to ‘coexisting’.
What does this mean? In the past, most people have seen humans and robots ‘collaborating’ on one task but there is a growing acceptance that, in most cases, we want robots and humans to be in the same space working side by side but not necessarily on the same task; this is ‘coexisting’. Co-existing gives us the opportunity to have a robot working on a production line next to an operator without the need for separation or complex collaborative safety cases (although some safety level will still be required). This frees up space and provides a lot of flexibility in the workspace in the same way that a human operator would but gives you all the benefits of a robot at the same time. It’s the approach that both P&G and Continental highlighted at this year’s European Robotics Forum with their ‘cobot’ installations. For both, the humans and robots do not do the same task, they each have their own ‘objectives’ but they exist in the same space and the robot reacts to what the human is doing to maintain their safety. The approach has proven really successful and one I think should be more widely adopted.
So where do we go from here? I firmly believe that humans and robots (of all sizes and speeds) will both collaborate and coexist in the factories of tomorrow. In certain conditions this can be achieved today. But for this to happen more widely there are some key fundamental steps that need to be taken:
- We have to understand the safety case for the whole application, not just the robot. This includes the end effector, any tooling, the process being carried out, etc.
- Design out any residual process or tooling risks
- Employ the right robot for the right job – if you need repeatability, fast cycle times or heavier payloads, a cobot probably isn’t the right robot but if you really want one, then understand its limitations and what you need to do to ensure it’s safe
- You can collaborate with a standard industrial robot, but understand the limitations of doing so (i.e. you won’t be able to go as fast, you’re robot cell might be bigger and the design/safety of the process will have to be considered or changed)
- Read the standards or seek help from someone who can help you
- Collaborate with research centres like the MTC; they need to understand what your needs are so they can develop the next generation of standards and technologies that really enable human robot coexistence.
Finally and most importantly, we just need to use more robots in UK manufacturing. Coexisting or collaborating robots are great, but for the majority of applications (both now and for at least the next decade) traditional industrial robots will be the better approach. But the UK is 22nd in the world ranking for employing industrial robots unless we change that, ‘cobots’ will never really give us the tangible benefits we expect them to. Some see an increased use of collaborative/coexisting robots as a way to increase our uptake of robotics and that’s fine up to a point. But unless we can encourage a wider use of robotics generally, the UK will still be missing out on the productivity, consistency and job creation benefits that robots are proven to give.
I am speaking at the forthcoming robotics and automation event at the IMechE: http://events.imeche.org/ViewEvent?e=6752