Key technologies defining smarter factories: the rise of cobots

This is bringing about more efficient production techniques, extending the operational lifespan of equipment and leading to heightened profitability

There are changes taking place on the factory floor too, in relation to the robotic technology employed. For decades large robotic manufacturing cells have been used for high-volume scenarios with fixed functions, enabling fully automated production lines that can function without any human intervention, in what are referred to as ‘lights off’ factories. These require a large quantity of supporting electronics hardware, as well as complex control software. Also, these robots must be kept within safety cages, to prevent them from injuring people through their movements.

A new type of industrial robotics is starting to emerge, though. One of the most exciting trends associated with the onset of Industry 4.0, this will allow other areas of production to be attended to. Collaborative robots (or cobots) are items of autonomous robotic machinery that can work alongside human operatives on production lines, often as part of a low-volume, high-mix strategy.

units may be trained by staff to replicate their movements without the need for complex programming. The objective is for cobots to complement the activities of the human workforce by performing the repetitive tasks, or taking care of heavy lifting, while their organic colleagues concentrate on the elements that make best use of their knowledge and experience (such as any fine-tuning that might be required).

The cobots’ software can store information on a multitude of different movements relating to a given manufacturing process, furnishing them with a high degree of flexibility. They conduct their work at a similar speed to human operatives, with no sudden movements that could prove dangerous. Furthermore, they are not as bulky as traditional factory robots. Once fully trained, cobots can run independently alongside staff without any risk of harm.

Ensuring coworker welfare

Safety is of course paramount for all robotic systems, but especially when human workers are going to be in close proximity and it is no longer possible to rely on safety cages. For this reason, an array of sophisticated sensor mechanisms (mainly machine-vision based) are integrated into cobots so they are fully aware of the position of their human coworkers.

Image sensors on the end of the robot arm follow the actuators to provide accurate location information, but also to identify any obstructions (whether passive or human). Proximity sensors can also be utilized for positioning the arm and avoiding obstructions, and the safety protocols in the controller can stop the arm quickly to prevent accidents.

Image analysis of the data from cameras has to be performed locally, to achieve the millisecond latency that an effective closed-loop feedback control algorithm calls for. This needs to encompass multiple axes of operation. Most cobot arms will handle four to six axes of movement, with different orientations of the gripper at the end of the arm.

This can represent a major challenge for a traditional image processing system, but through assisted learning (where the arm is positioned by the operator) the accurate images that a cobot requires can be provided. Consequently, cobots can continue to keep learning on the job, and this in turn helps to improve the accuracy of the image recognition.

Sensors within the arm also monitor the condition of the motors, feeding back data to the relevant staff (so they are aware of any potential issues requiring maintenance). Cobots are also quick to deploy, which means they can heighten the productivity of an Industry 4.0 facility in just a few days (rather than weeks). Working alongside highly skilled staff, problems can be easily identified and fixed at an early stage, and output volumes then ramped up accordingly. 

Cobot uptake and evolution

Examples of the effectiveness of cobots are being seen across a broad cross-section of industry sectors. For instance, a Spanish cosmetics company has been using six cobot arms in its packaging plant to stack boxes on pallets. These units can support a stacking rate of six packages per minute, and can deal with a portfolio of 350 different product types. Similarly, a Taiwanese injection molding firm has added four cobot arms to its production line, in order to tackle a serious shortage of skilled staff.

 Following training, these units have been able to cope with the high-mix requirements of this industrial process and offer the degree of product customization needed.

The ability to grip sensitive components accurately (without causing damage) will significantly increase the range of applications that cobots can address. As a result, there is a lot of research being undertaken into new types of actuators for incorporation into cobot arms. Scientists at the University of Buffalo in the US have developed a two-fingered magnetic gripper specifically for cobot use. Its innovative design absorbs impact energy, preventing breakages.

Instead of having two fixed fingers, the gripper has a magnetic base that sits between two neodymium magnets. These repel each other to create an air gap, and this acts like a buffer against shocks. The stiffness of the grip can be adjusted instantaneously, by simply increasing or decreasing the space between magnets. Feedback can be taken from the sensors to provide the right level of grip.

In one test undertaken by the scientists, they managed to get the gripper to hold a length of spaghetti between its fingers. When the gripper came into contact with this fixed object, the sensors detected the external force. Using the feedback given, the magnets were adjusted to reduce the stiffness of the grip, and the spaghetti remained in one piece.

Conclusion

Cobots are showing significant potential for driving the digitization of our factories, bringing automation to production lines that would not previously have been possible. Another advantage is that cobots can be rolled out quickly onto production lines that are already in operation so as to automate existing processes, rather than having to redesign an entire production line from scratch.

Not only do they monitor themselves, they also feed data back to central servers to be part of a comprehensive digital model of the factory (the subject discussed in our next blog). All of this provides the detailed visibility and granular control that is going to prove essential to Industry 4.0.

What will this blog series cover?

1. Key technologies defining smarter factories – connectivity
2. Key technologies defining smarter factories – sensors
3. Key technologies defining smarter factories – the rise of cobots
4. Key technologies defining smarter factories – digital twinning
5. Key technologies defining smarter factories – AI
6. Key technologies defining smarter factories – data security