Insights accelerate innovation in digital manufacturing
By Andrew Quinn, Digital Advisor, and Benjamin Wright-Jones, Architect, Data and AI, United Kingdom on February 14, 2018
Filed under Discrete Manufacturing
Optimizing product development, improving speed to market, fostering employee productivity, and driving a data-driven culture.
Our team at Microsoft has been collaborating with a Fast-Moving Consumer Goods (FMCG) manufacturer to tackle these business problems within R&D early-stage manufacturing. We have enabled digital manufacturing to deliver data insights through new experiences that also support changes in the company culture.
Together, we have enabled disruptive innovation through sensor telemetry, observational, and system data while optimizing manufacturing processes and improving product quality. These provided us with deeper insights and analyses.
The Internet of Things (IoT), which is an evolution of Machine to Machine (M2M) communication, was at the center of our approach. Because of the analytics and artificial intelligence that can be deployed, IoT opens new business opportunities not previously available with M2M. Companies typically ingest and process data from IoT-enabled devices into a cloud platform, potentially enrich it with additional data such as human observations or offline measurements, and then act upon it or use analytics to gain insights in near-real time. As the outputs, insights, and potentially, recommendations, are fed into existing systems, they drive innovation.
We posed the question: Could we use IoT and innovation opportunities to improve manufacturing?
Throughout our engagement with the FMCG manufacturing company, we fostered close collaboration between designers, developers, data analysts and the process engineers and operators to iterate on requirements in an agile manner. This mutual learning and level-setting of knowledge increased their comfort level, which meant we could more easily discover real user needs, and later, improve the likelihood that the users would accept the tools in production. We also drove an open data culture, enabling the organisation to share information more broadly between manufacturing sites and across the supply chain.
A core tenet of ours was to deliver innovation through a true digital manufacturing platform, and not just digitise existing ways of working. This project would provide operators and engineers with new insights from machine telemetry and human observations. Our agile approach would ensure flexibility, help us meet deadlines, and reduce unnecessary development efforts, and related costs.
We started by developing a digital manufacturing solution which would first address user issues, and then allow us to continue improving the underpinning service capabilities. Users were provided with Power BI visualisations and a tablet-based application that provided insights from multiple sources such as manufacturing equipment, SAP systems, and human observations (audio, image, text).
We developed leading-edge configuration management and release management strategies to address challenges such as how to incorporate, or manage the removal of, a sensor from a manufacturing vessel or rig, how to manage sensor locations and taxonomy (friendly names), and how to maintain application code, apply updates, and keep code bases in synch when new hardware devices were deployed. These challenges were particularly complex because of the number of hardware and software vendors involved.
We worked with hardware manufacturers such as Siemens to understand how their hardware could be network-connected to extract data from sensors. We discussed data security (authentication, authorization, at-rest, in-transit), and industry compliance and technology capabilities, all while considering the company’s unique security controls – including network monitoring and firewall management. We found that sensor-generated data had to be decoded by experts that could translate the values; this required involvement from manufacturing domain experts, equipment specialists, and developers.
Even though an open data culture would let us combine digital manufacturing data and knowledge with external partners and potentially, academic institutions, it also made data governance trickier; we had to consider data stewardship, traceability, ownership and accountability.
In addition to handling these technical aspects, we supported and educated business owners as they identified the data to capture, how to capture it, and how to use it for maximum value. Business and IT department stakeholders participated early in envisioning, solution definition, delivery and deployment to ensure they’d use the systems when they were launched. We spent a lot of effort ensuing a strong working relationship between the development team, business leaders and the manufacturing process operators and engineers. Office co-location facilitated the culture – supporting the agile way of working, enabling a fail-fast mentality and strong team collaboration that in turn, enabled rapid and iterative feedback during sprints.
In the end, we were successful through our combined culture and approach. We had an empowered stakeholder that took the responsibility for user story definition and prioritisation, we had a senior business stakeholder that appreciated the value of agile methodology / way of working, and challenged their own ways of working to adapt their business to the technology. Prioritization was key because it was easy to get distracted with the many possibilities of IoT. Lastly, we shared the perspective of not fearing failure- allowing everyone to experiment and learn.
Learn more at Microsoft Digital Advisory Services.