Adaptable robot adopts machine vision

A major building products supplier is among the first to adopt the new generation of vision enabled robots.

A major building products supplier is among the first to adopt the new generation of vision enabled robots.

Able to make decisions based on sensory feedback from their environment, vision enabled robots can adapt to perhaps any manufacturing scenario.

Robots traditionally were taken up more by bigger manufacturing outfits which could invest in both the technology, and a supporting infrastructure.

But as costs continue to fall, the technology itself is changing in ways that mean robotics can benefit almost any manufacturing environment.

As well as advances in robot mechanics and developments in their controllers, the integration of robots with sensors is powering a new future for new and traditional users alike.

Robots are routinely rigged to incorporate proximity, touch or laser sensors, but recent advances in the use of machine vision (MV) enabled within robots is causing a stir.

Historically, the benefit of robotics is high repeatability, however if the product isn't repeatable, or in a repeatable position, then additional infrastructure is needed.

Take the typical function of palletising.

To enable a robot to pick something off a conveyor, the manufacturer would have to construct a fixture to hold the product in a repeatable position.

Changing the product would require changing the fixture, or some cases, the fixture itself could be overly complicated to construct.

For one major building products supplier using vision enabled robotics (VER) to palletise concrete paving slabs, the benefit is clear.

The company has 42 product variants: sizes, colours and shapes varying from squares and rectangles to quadrants, circles and ellipses.

The mechanical solution for a square slab is to have four paddles push in to centre it, but this is not easy for a circle, a triangle or a rectangle.

Instead of fixtures, a vision system uses two slab position checks on the conveyor prior to palletising.

The VER accurately determines the centre and the rotation of each slab so as to correctly pick it up and place accurately to the pallet, prior to banding and shrink-wrapping.

No fixture is required and the slabs can be in any position on the conveyor.

Such functionality may be commonplace nowadays.

But the supplier is using VER: not a robot working alongside a smart camera, but one with its own machine vision (MV) "enabled".

In this setup, a DVT Corp vision system is directly connected to a Kuka Robot controller via fast Ethernet.

MV is no longer managed separately from the robot, nor does it require its own PC monitor.

Rather, MV is accessed directly from the robot's teach pendant, a handheld screen with simplified input buttons.

A cost benefit is derived from lightening fast data transfer speeds.

As the concrete slab moves into position, the VER takes three images, compares them and calculates the mean location.

In the old method using a separate camera and robot, the transfer of this information could have a dwell of perhaps a second, while the two systems analysed data.

Now the total time to palletise the product is just 4.5s.

The time saved, multiplied per hour, per day, per year, means more products can be palletised - in effect for free.

"This reduces cost and simplifies operation of the robot's vision", says Ian Tatton, DVT's Director of Business Development for Northern Europe.

"It also means the operator's perception is of a robot with integral vision, so there is going to be an increased expectation that the robot will perform tasks that incorporate a vision function".

Brett Green, General Sales Manager at Kuka Robot, sees the overall benefits as much to do with added value as cost and time savings.

"We also use the image to determine if there are any corners missing, or if there is any other defect".

"If detected, the product is scrapped and not palletised, so QA improves dramatically".

"There are also benefits for distribution, the builders merchant and garden centre receive uniform packaged product, and once palletised and no one sees or touches the product again".

The heart of the VER's ground-breaking technology is its twin GUI, a joint development between Kuka Robot and DVT Corp.

This changes how robotic vision is managed and operated.

DVT's Framework MV software remains accessible through the robot's teach pendant, so a trained operator can set up any number of work routines, inspections, tests, locating co-ordinates, scans, OCR activities or other tasks.

Kuka Vision is the new second tier user interface that utilises DVT's ActiveX control to provide a simplified, operation-only version of the software.

This means that while one trained operator can program a variety of tasks, any operator can perform them without training.

Kuka Vision provides the operator with intuitive one-click controls governing key functions of the robot's vision system, including tasks such as: connect to cam; see or record an image; see data; or switch the robot's product routine.

With just one trained MV specialist, a variety of complex routines can be created, so operators need not be trained in depth, and the robot is increasingly autonomous in any case.

The total cost of ownership falls, while even the MV specialist has ongoing DVT training free of charge, to keep up with the rapid developments in machine vision technology.

As complex routines become programmed into the robot, its enabled vision can be taught to recognise different products that trigger different routines.

Brett Green, General Sales Manager at Kuka Robot, explained: "By being able to interact with its vision system (effectively giving the robot a pair of eyes) it allows the robot to change its programs to accommodate what its seeing".

In the case illustrated here, different slabs are inspected and stacked in particular ways, just as in another scenario, a robot might use its vision learn paths to locate a variety of tools that can be used to perform different tasks.

With cost and time savings, operational improvements and the ability to carry out new, more efficient functions, improving added value and achieving 100% QA in many cases, what of the future?.

Brett Green of Kuka is definite: "We will see a lot more sensor robotics, meaning they will have an improved ability to adapt and respond to their environment and the task in hand".

Ian Tatton of DVT said: "Combining with partners such as Kuka, different perspectives and expertise is opening up new applications for machine vision in robotics and other areas".

"This helps us to achieve our aim of smaller, faster, more powerful and less expensive machine vision for all".

Back to top