Trends in AI and robotics beyond 2020: Four key challenges and opportunities
08 January 2020
With 2020 well under way, many companies will be looking to identify the key trends and developments over the coming year.
Artificial intelligence (AI) has been the subject of considerable hype for several years. but the industry is now ready to move to the next stage. It’s now time for the emerging practical industrial automation solutions to focus on ways in which AI can enhance human intelligence, or augmented intelligence, and how it can be implemented on the factory floor, whilst ensuring sustainability. These solutions will support people working in industrial automation in decision making and improve operational efficiency.
Here, Omron have identified four AI trends in robotics and industrial automation to commence the start of the new decade:
Valuable machine data generated at the edge
A new generation of workers in industrial automation will change jobs more frequently than ever before. The latest developments in factories depend on the generation and collection of deep knowledge and data insights at machine level – i.e. at the edge. The machine can learn from its human operators and subsequently improve the output. Technology controlled by AI can empower machine learning by predicting both product and equipment failure, using data generated by Industrial Internet of Things (IIoT) devices. The analysis and use of combined data enable users to rapidly predict potential machine errors, preventing disruptions and the deterioration of product quality.
Increased efficiency through self-learning algorithms
With the change from mass customisation to a high-mix, low-volume approach, efficiency must be improved by reducing human errors and machine downtime. AI with learning algorithms can help machine operators to achieve the best result in every change-over. Innovative control technology can also help employees to work alongside robots and machines to achieve manufacturing excellence. This is accomplished by using a broad range of factory automation equipment that enables IIoT-capable production or implements optimal AI algorithms in the equipment. The AI-equipped controller is designed to immediately detect signs of any equipment irregularity. The machine automation controller's AI algorithms allow it to learn the repeated movements of equipment from precise data from sensors. This in turn provides feedback for status monitoring and the real-time control of machines.
Efficient decision-making with visualised data
Industry 4.0 and IIoT enable the accurate collection of historical data. However, many AI projects struggle with the visualisation of new data. Predictive maintenance and control solutions can align the control functions of manufacturing lines and equipment with AI processing in real time. They can support companies by generating new, rather than historic, data that is time-stamped and easy to visualise. The process of collecting raw data from machines is completely automated, using an AI controller which operates on the ‘edge’ within the machine. This leads to higher data accuracy and consistency. A controller can also create data models from correlation analysis and monitors the machine status based on these models. Without this automation, machine designers and operators would need to invest in developing their own analytical and optimisation capabilities.
Sustainable technology
As the world’s population continues to grow, this places an increasing burden on the environment. AI-assisted collaborative robots (cobots) will play an increasingly important role beyond 2020. The aim is to create healthy and safe living and working conditions that cause less harm to the environment. Omron can help companies to achieve more sustainable working conditions in factories with its portfolio of robots and AI. Assembly and disassembly robots have an important role to play here. The new generation of robots can learn from machine operators (sensing) and collaborate with cobots (control) on a circular production line. They collect smart and intuitive data from its actions, assess the data using algorithms, advise the operator about the next steps, and implement efficient processes for each changeover (think).
Contact Details and Archive...