Data Analytics For Dynamic Manufacture
Current limitation in industry
In 2017 the UK Government announced that Artificial Intelligence (AI) would be a core focus of its Industrial Strategy, elevating ‘Big Data’ and machine learning to become one of the flagship technologies that will transform the future of the UK economy. It is clear that the digital revolution is underway. While company leaders are figuring out how to utilise this to gain competitive advantage at a global level, cybersecurity remains a critical issue that impacts organisations’ uptake and integration of data into existing business processes, especially for companies within which automated data-driven innovation is not traditionally associated with product and service improvement. This includes the UK manufacturing sector where there is a crucial safety element to consider, but which, according to a 2016 report from the UK Office of National Statistics, could be massively boosted with these technologies. The report stated that “Britain’s hourly manufacturing output is approximately 27% below France and Germany, and 31% Below the USA”.
Current limitation in scope of research
The Internet of Things (IoT) also has an increasingly significant role in this sector. With the ability to continuously collect and relay data to a central hub from distributed sensors. The combination of embedded sensors (which are becoming increasingly smaller) with AI has the potential to transform the product manufacturing process, radically interrupting the existing ‘consumer sovereignty’ model based around surveys and market research – and introducing ‘use sovereignty’ via an embedded understanding of consumer behaviour observed via IoT. The manufacturing industry can then take huge leaps towards making products that are fit for purpose based on how they are used in a quicker and more efficient manner.
Extensive research has been conducted into cyber risks within organisations but there is a lack of evidence on the risks faced by converging emerging technologies such as the Internet of Things (IoT) and AI in operational settings – such as Operational Technologies (OT) on the factory floor. We will study the implications of opening up the factory floor for direct communication between emerging technologies and OT. This will not be unmediated but buffered by robust, secure and interpretable data analysis at scale – with the ethical integration of human labour. Designers will have a completely transformed role being ’embedded in production’ in a world of “chatty” products and a dynamically evolving factory floor.
The Value added to research
A notable value-add to academia is the novelty of contributing to the body of literature on cyber security and AI, with a focus on secure and perimeterless operational and information technologies within the manufacturing sphere. In addition, empirical research will evaluate the risks and develop security maturity model for opening up the factory floor to free-flowing product use data.
The Value added to industry
A key component of the transformation plan, will be to strategically collect and analyse large volumes of data from sensors embedded in products; these product sensors will send valuable, real-time information back to the manufacturer to enable dynamic product analysis and inform real-time product re-design – which will ultimately be dynamically pushed to the factory floor, creating an endless production-consumption feedback loop. For instance, we will be able to understand the trends of the product ‘experience’ across the whole consumer market. If a product is dropped and cracks in the same place for 90% of customers, the product design can be updated to reflect this, saving recalls or undetected safety issues.
As part of our research on data analytics for dynamic manufacture, our initial value-add propositions are as follows:
•Designing an effective data collection process within the factory and beyond through the deployment of micro sensors inside “chatty” products.
•Understanding the practical implications of opening up the factory floor by investigating new data security requirements (GDPR) and the risks of converging Information Technology (IT) with Operational Technology (OT) on the factory floor.
•Investigating novel computational intelligence and mathematical methods to reduce noise in product sensor data and identifying ‘useful’ data to inform an understanding of how products are used and how to optimise them.
•Investigating and developing an auditable data-driven manufacturing process model.
•Developing new methods to capture provenance of design and manufacturing changes.
Visualising the future
With “Chatty Factories”, we see a future that empowers UK manufacturing companies with the capability to harness AI and ‘Big Data’ (generated from sensors embedded in various types of products) in real-time, integrating this as part of the design and manufacturing process by feeding information that reshapes not only the manufacture but also training the robots and humans on the factory floor – ultimately cutting long-term R&D costs and optimising the production process all wrapped with an intuitive and adaptive IT/ OT Security Architecture.