Professor Samia Nefti-Meziani takes a look at robotics, soft cobots and cognitive automation and how they offer an overarching approach towards a human-centred smart production environment.
By Professor Samia Nefti-Meziani
For UK manufacturing to remain competitive, it is imperative to focus on delivering products efficiently in a cost-effective way without sacrificing quality.
Most of the quality control and metrology solutions used by businesses to provide accurate and reliable measurements of products has, until now, relied heavily on hand-held tools and worker's dexterity. This is a core issue as the overall performance of the process could be compromised by human error, especially when tasks are performed repeatedly and inconsistently. Moreover, there are some other solutions available for metrology-assisted operations, including fully automated machining cells incorporating metrology tools.
A more interesting metrology solution that may be cost-effective for SMEs and could ensure manufacturers' production performs better is the semi-automated system. It can offer an autonomous parts handling system with integrated inspection but requires a high human intervention level. This can be achieved by solving manipulation problems, automating the absolute measurement and higher repeatability in inspection tasks, to make it an integrated part of the production process. This is where collaborative robots, i.e. cobots, come into play. While current conventional industrial robots prove to be highly effective on the assembly line, they remain heavy and isolated from the humans on the factory floor because they can be dangerous to be around. In contrast, Cobots emerge as a solution to improve the task execution where the human is required.
Cobots can provide the necessary force, repetition and accuracy required for tasks, including polishing and measuring. They can be taught manually using Artificial Intelligence to help humans ensure efficiencies and maintain the quality aspects of the product, significantly so when the process’ overall performance is compromised by human error. On the other hand, cobots are a smaller size and easy to move across the shop floor without changing the production layout. They are easy to assemble, take apart and re-task. Their joints are force limited, making them very safe in case of a collision with a person – although their safety approval could be difficult to obtain, especially if the end effectors used present some safety issues.
However, the existing cobots also have serious limitations, including the low payloads that they are able to handle (between 3 to 15kg), and their typical speed is roughly considerably less than a traditional industrial robot. In terms of precision, industrial robots can produce faster and more accurate paths for a task than cobots can do, as it is challenging for workers to generate an accurate path of the task's motion that the cobot would need to perform.
In contrast, the new generation of cobots, i.e. soft cobot, with their unconventional elastic materials, take advantage of deformable materials and their intrinsic dynamics to enhance flexibility and controllability. In so doing, this makes them substantially more human-friendly and could potentially overcome the limitations in speed, handling, lifting payload, cost and the accuracy of the existing cobots.
However, for soft cobots to be deployed for quality control and metrology tasks, additional steps toward providing a credible alternative to traditional rigid-bodied robots and existing cobots are required. There are still some technical challenges that remain to be addressed. For example:
- the ability to modulate stiffness in soft robots is a prerequisite to suit specific tasks (especially when accuracy is required), and despite substantial research, the soft mechanisms proposed to date are still limited and need to be developed further;
- soft robots are continuously flexible/deformable. The need to cope with different manufacturing environmental conditions, including humidity, temperature, vibration, and cleanliness, can be constraining factors that mean traditional sensors are inappropriate and advances are needed;
- soft robots have demonstrated superior manipulation in some applications compared to conventional systems, but universal soft manipulators capable of handling all products types are still largely missing;
- there is also a need for control and architectures that exploit morphological computation and control methods to external and system damage.
Addressing these challenges will allow soft cobots to become more compact in size, more accurate, more dexterous, easier to control, and more affordable.
Professor's Samia Nefti-Meziani research group has already tackled some of the above challenges by pioneering the first application of this new generation of cobots in manufacturing. Her team has been addressing these challenges since 1990 through prototypes and demonstrators, leading on resilient, soft robots.
For example, Professor Samia Nefti-Meziani and Professor Steve Davis recently demonstrated a 40cm soft arm able to lift a weight of 150kg, higher than is typical of soft systems. It addressed the stiffness modulation challenge, advancing soft sensing through the development of a soft skin and integrating sensors into soft actuators. They also developed the first example of a pneumatic muscle able to self-heal and have been active in developing universal soft end effectors for handling all types of objects and payloads.
By addressing the above challenges and increasing the UK's research capacity in this area, we can expect true breakthroughs with enormous importance in the short term in manufacturing. Soft cobots can provide cost-effective technology for SMEs and businesses to increase their productivity, reduce waste and ensure that parts used in aircraft, cars, and trains, for example, can be made more lightweight. This in turn decreases their energy requirements and lowers the amount of CO2 released during operation.
Manufacturers will undoubtedly benefit from research in soft cobots that will refine the collaborative interactions between humans and offer low-cost solutions for metrology and inspection tasks. Soft cobots will also present some interesting challenges to the metrology community in developing sensors and measurements techniques. Industry 4.0 has clearly made an impact on manufacturing, but if we want to take deployment, safety, quality, productivity and reliability a step further, we need to look ahead, embrace the next generation of cobots and pave the way to Industry 5.0.
For further information please see: