Vision systems and robots move closer


UKIVA Chairman, Neil Sandhu, says the increasing convergence of vision systems and robots will create a new generation of industrial equipment that combines their unique advantages.

Since robots and machine vision systems were first industrialised in the 1950s, the two have developed in parallel. Both are now used routinely in manufacturing. Over time, they have also begun to converge – developing something close to a symbiotic relationship.

Combined robot/vision systems are probably best known within assembly operations. However, they can be used across many areas of manufacturing – including inspection, sorting and quality control. A number were seen at UKIVA’s Machine Vision Conference (MVC) earlier this year. Examples of successful robot/vision systems included inspecting electric vehicle batteries, improving the robotic harvesting of crops, and a number of pick-and-place applications.

In another application, an automated cell used a collaborative robot (cobot) to weld a number of bosses onto a cylinder. This required a vision system that could achieve several tasks: classifying the cylinder type, locating the bosses, and measuring their diameters and positions. It also had to be accurate to 0.5 mm.

The cobot used profile sensors from Wenglor to find all the necessary information – which then allowed a welding path to be calculated. Each sensor contains one ‘recipe’ for all cylinder sizes – and has no need for additional lighting. It’s a perfect illustration of how the whole is greater than the sum of its parts. It works thanks to the seamless combination of robot and vision system.

Visitors to the MVC would have seen several exhibitors – all of them vision companies – with some form of robot on their stand. This illustrates the high degree of crossover between the disciplines. The overlap continues to rise: it’s fair to say that most people looking to use a robot these days will also require a vision system to go with it. Precise robot guidance requires an accurate vision system.

There have been huge advances in interlinking – that is, getting a robot and a vision system to talk to each another. This is now much easier to achieve, thanks to the existence of common operating protocols – driven in part by increased collaborative research. In addition, there is no need to be a programming expert, as a lot of the work has already been done. These vision/robot systems are increasingly plug-and-play. This makes deploying a complete system more straightforward.

Training a vision system to recognise multiple objects using deep learning can enhance the performance of a robotic system. A good example of this – also revealed at MVC – is an automated waste-sorting system developed by UK company RecyclEye.

The RecyclEye system uses RGB sensors and a sophisticated algorithm to separate waste at high speed. It is more effective than traditional methods such as near-infrared sensing or manual sorting. For instance, it can determine food-grade plastic (such as milk bottles) from non-food-grade material (detergent bottles) with 96% accuracy. To do this, it relies on a huge dataset of scanned images.

Having been tested in recycling facilities – a highly challenging environment – the company believes the system is robust enough to be applied to other sectors.

The intertwining of robotics and vision systems is set to continue – and this has informed plans for a new event. MVC 2023 will become a two-day event and run beside Automation UK – a new show focused on robotics organised by UKIVA’s sister organisation, BARA.

The co-located events will run at the CBS Arena in Coventry on 20-21 June 2023. MVC will remain a conference with an attached exhibition, while Automation UK is intended to be a national automation and robotics show. Together, they will address the rising convergence of robotics and vision technologies – through both presentations and exhibits.

Robots and vision systems have become increasingly inseparable. In addition, both have become more affordable, capable and accessible to users. In most cases, a vision system empowers a robot to raise its performance.

We’re looking forward to proving this next year by helping manufacturers understand – and deploy – the best robot/vision combination for their applications.


About Author

Comments are closed.