The future of the robotics in the manufacturing sector
Robotics in large manufacturing companies has been widely deployed across a variety of tasks. This includes material handling (73%), welding (4%), assembly (drilling, fastening, fitting, riveting – 2.4%), processing (gluing, painting, polishing, routing – 1.2%), packaging (<1%) and inspection (including clean room – 9.4%) [IFRIR17]. Take-up of robotics by Australian organisations, however, is producing low volumes, and customised or complex products have been limited due to inflexibility and the high barriers of entry to existing robot systems.
The opportunity to apply new robotic technology in the manufacturing sector is two-fold: to enable seamless and safe co-operation between robots and people, and to allow rapid adaptability of robots to new tasks without requiring a deep automation skillset on behalf of the manufacturer. Put simply, enabling a new class of robots to think and see has the potential to drive a step change in Australia’s manufacturing competitiveness and productivity.
Adding a robot to a production system yields improved quality, safety, and productivity. The productivity of person-robot teams is greater than either person teams or robot teams alone. The deployment of a collaborative ‘assistant’ robot that can communicate and work safely with people in manufacturing environments would unlock latent productivity and augment Australia’s world-class talent pool.
SMEs typically lack the resources to have dedicated automation engineers on staff. This limits the uptake of robots into low volume, or highly customized, production lines, as the programming cost can overwhelm the productivity improvements. Using AI and vision to build robots that can learn skills from experience or demonstration, and share those skills across industries or organisations, will further allow robots to be added to Australian factories at low cost and without requiring highly skilled programmers to be on staff.
Low volume, custom manufacturing relies on high quality to be cost efficient. Without the repetition and structure of a high volume standardised manufacturing line, “blind” robots can cause defects or pass on incomplete work. Equipping robots with integrated vision and other sensors enhances the information gathered during the manufacturing process. This allows robots to assess the job to be done, make any required modifications to ensure process quality, and record the outcome for auditing.
SMEs may not be able to easily quantify the impact of adding robotics to their manufacturing lines. High fidelity simulation enabling virtual prototyping, design, and commissioning of automation will allow manufacturers to reduce risk. Benefits can be validated before significant money is invested in hardware, and improved speed of deployment can be achieved once the design is finalised.
Achieving this step change is not a technology challenge alone. The Australian robotics supply chain must be equipped to integrate and supply these next-generation solutions to manufacturers. A parallel effort to develop shared libraries and tools to enable, and encourage, IP leverage across industry sectors will allow capability to be used by Australian manufacturers and be shared with other Australian companies. Benefits to Australian companies include:
- Rapidly adding a new production line to a factory, without manual programming of process steps, logistics, layout, and without explicit expertise.
- Enabling a future where assistant robots routinely work cooperatively with people in semi-structured manufacturing environments, communicating using natural language, and explaining decisions.
- Building sensor networks to provide live data as a service to enable holistic quality, logistics, safety, and robotic function while integrating into manufacturing systems and decision-making tools to continuously self-improve operations.
- High fidelity simulation allowing virtual prototyping, design, and commissioning of automation to reduce risk and validate benefits before significant money is invested in hardware.
- Having solutions that can be quickly and easily adapted to different tasks and industries.