PARC’s chief executive Tolga Kurtoglu explains how Internet of Things-based data analytics brought maintenance efficiencies to the East Japan Railway, despite the challenges of ageing infrastructure and workforce
“What we are seeing today is challenges resulting from global megatrends related to the pace of technological change,” says Tolga Kurtoglu, chief executive of the legendary IT and hardware developer PARC – the Palo Alto Research Center.
Since the 1970s PARC has gained a reputation for delivering innovations that now form part of our everyday lives, including the laser printer and computer mouse. Today, from its base in California, PARC is working on cyber-physical systems, human-machine collaboration and sensing.
Kurtoglu says: “Our job at PARC is to understand the business problems that arise from current challenges, and then to apply emerging science and technology.” By way of example, he refers to one of PARC’s recent applications, in which improvements in maintenance efficiency and safety on the East Japan Railway have been achieved by using “new ways of sensing, analytics and machine learning, related to the Industrial Internet of Things”.
The old way of doing this was with schedule-based maintenance (SBM). Kurtoglu says that PARC is assisting in shifting the focus away from SBM to ‘condition-based maintenance,’ which employs analytics to monitor in real time the health of a system, to understand its usage patterns. “We’re seeing the emergence of computing platforms in data analytics that are supporting a paradigm shift in how we approach maintenance.”
Kurtoglu says PARC developed the realtime data analytics platform MOXI “to enable predictive condition maintenance. It’s a combination of machine-learning techniques and first-principle model-based techniques to form a holistic platform where we can look at and understand the physics and dynamics of complex systems. We then tie that into the ability to make predictions about abnormal system behaviour.”
Back on the East Japan Railway, there were two reasons for moving to MOXI. First was the desire simply to be up-to-date in response to ageing infrastructure. But there was also the need to preserve institutional memory in the context of an ageing workforce.
It was necessary to counter these challenges, while servicing the trains in a cost-efficient manner, with increased safety and less downtime. “How could we bring about digitalisation in a way where we could capture the outgoing expertise in a computer-interpretable fashion?” Kurtoglu adds: “We look at the areas requiring high-impact solutions and build platform-based capabilities to enable digital transformation. This is the model for managing change. That’s why we developed MOXI.
“But it can be applied across many vertical sectors that have the same characteristics of ageing infrastructure and a workforce approaching retirement. We see this not just in transport systems, but also in infrastructure applications such as the power supply grid.”
Interpreting data is only one part of the new problem-solving model. “I want also to emphasise the physics, by which I mean sensing. Often in Industrial Internet of Things systems it’s not about gathering a lot of data to get you to the solution. It’s about getting the right data. And, if you can’t get the right data through existing sensing modalities, you need to invest in new technology.”
It is also about preserving in the workforce the institutional memory that will create continuity of expertise. “The technology in itself won’t do that,” he concludes. “We need to deliver the solutions in a way that provides seamless transitions from generation to generation.”
Article Source: IMechE