Subscribe to our Newsletter | To Post On IoT Central, Click here


predictive maintenance; industrial; iot (1)

We are in the midst of the fourth industrial revolution within the Industrial Manufacturing sector. Accenture & General Electric recently published a paper on the Industrial Internet that places spending at $500 billion by 2020 and forecasts growth to a whopping $15 trillion of global GDP by 2030. The Industrial Internet is disrupting almost all industries today with a revolution of possibilities and opportunities across the value chain. The third Industrial revolution ushered in the digital era, but Industry 4.0 has the vision to connect the digital to the analog world in a truly seamless fashion.

Overcoming the challenges of systems architectures of the past

As industrial systems grow larger, their architecture frequently needs to be updated to match the technology capabilities of today. Most industrial systems have been closed systems built on proprietary technology stacks, both the hardware and software. It was impossible to communicate with them: communication worked only within the systems’ own ecosystems. Interoperability has also been challenging, and at times nonexistent, due to incompatible interfaces even among parts of the same family of systems. These systems could analyze real time data and use historical data to make good decisions, but many were built with rigid interfaces that lacked the ability to exchange data with others outside their ecosystem.

Where will we go with Industry 4.0?

As physical and digital worlds come together via exponential growth in analog and digital integrations, those who implement manufacturing control systems will need to understand the complexities of Industry 4.0 and lead the way to simplify them. Most industrial solutions now have embedded control systems that are constantly monitoring and computing based on feedback, so managers can optimize the performance of manufacturing processes. The next step in the evolution is to converge physical and digital processes and data to create a more holistic view not only of the manufacturing process, but also the context in which it operates. Both heavily impact realized outcomes.

The ‘things’ are absolutely vital for the Industrial Internet of Things, but what makes them highly valuable and ‘smart’ is when they provide unique insights and context and inform action by communicating, cooperating and collaborating with each other. Each ‘thing’ works not on its own, but rather in a mesh with others to attain a singular purpose. This also makes ‘things’ more resilient to faults and outages. They can be tuned to use little communication overhead and work in a congested and constrained network. However, an infrastructure to support communication among ‘things’ and addressability over networks and channels must be established via an Integrator or a Gateway.

A key component of Industry 4.0 is the Smart Factory, which should be context aware to help people and machines understand the execution context of the task at hand. This is different from the feedback loop of control systems. Rather, in a Smart Factory, the machine knows the state of operation not just by the position, but also from data that is provided to it from other information sources. This helps workers on the factory floor focus on higher priority tasks within context. For example, if in a warehouse, the sorting system needs to be calibrated after every couple of days, the system will automatically initiate an auto-calibration when it knows that it has no pending activities. The calibration data could be derived from the manufacturer or a local system that maintains it.

Use Case: Applying Predictive Maintenance on Heavy-Duty Gas Turbines

The energy sector offers a good use case for Industrial IoT, and particularly with gas turbines. The primary purpose of a turbine is to generate power. The turbine itself is made up of rotors, blades, exhausts, inlets, brushes, shafts and a variety of control systems that manage fuel injection, power generation, etc. Each of these turbine elements has to be maintained from time to time, usually on a scheduled maintenance cycle. Any scheduled downtime of the turbine has to be managed; otherwise, it is not generating power. By combining data from the array of ‘things’ that can monitor the various parts of the turbine, the frequency of vibration of the blades and rotors, tensile and/or radial stress, and leakage control at the various seals, turbine operators gain insight into the overall turbine health and not just selected aspects. In addition, adding data on environmental conditions, such as temperature fluctuations, humidity, air quality, geography, and fuel quality, provides valuable context on the running condition of a turbine. For example, data on air quality can be used to predict when air filters need to be cleaned or changed. 

Turbine operators can collect and store data from each of the fleet’s turbines in an Asset Performance System. By applying statistical models and looking for patterns in the data, turbine operators can optimize maintenance scheduling and identify common fault areas in order to take corrective and/or preventive actions before an issue occurs. This results in lower maintenance costs and less turbine downtime, so turbine companies can generate more revenue, increase profitability, and deliver a better customer experience.

The impact of the Industrial Internet will be far greater and widespread than other industrial revolutions before it. Companies are beginning to realize the financial benefits and early mover advantage from implementing Industrial IoT, and we have only just started to scratch the surface of possibilities.

Read more…

Upcoming IoT Events

More IoT News

IoT Career Opportunities