Companies are implementing Proof of Value (PoV) projects to validate Industry 4.0 and IIoT initiatives with the goal of demonstrating the viability of these technologies and understanding potential challenges and limitations to plant-wide deployment.
Most of these pilots fail to scale.
There are many reasons, of course—not having a clear business case, not involving the right people, problems with wireless connectivity and integration with other plant systems. One very practical challenge is the inability to monitor all assets (rotating and static). Even within rotating equipment, variable-speed assets present unique challenges. If asset selection is limited to only fixed-speed rotating machines, the value for the IIoT pilot may not address some of the more critical production assets.
Variable-speed assets
Variable-speed assets are used in applications across all industries. Any time there is a designed variation in the process (eg. producing different finished products), variable-speed assets will be more efficient compared to fixed speed. Typical applications include pumps, fans, agitators, cranes, machine tools, HVAC and food-processing machines. Speed is controlled by varying the voltage to the motor, providing only the power required for the application.
By reducing the motor speed, variable-frequency drives (VFD) offer savings of up to 75% of the energy consumption, which can be a significant over the life of the equipment. VFD equipment is still a small percentage of most industrial assets, but accounts for approximately $24 billion, according to Grandview Research, and is the fastest growing segment of the drives market. The growth has been driven largely by energy efficiency, operational performance and lower maintenance costs.
Challenges monitoring variable-speed assets
Condition-monitoring technology has been in use for more than 40 years and has traditionally relied on highly skilled, experienced analysts to periodically collect data using technology like vibration analysis.
This periodic, walk-around approach to condition monitoring requires consistency in data collection to allow trending over time, making it difficult to adopt on variable-speed or spared assets. Equipment is often not operating under the same conditions when data is collected, and changes in load or speed produces data cannot be trended (as the vibration levels will vary).
Without trendability, it is difficult to accurately predict when an asset is starting to fail. Similarly, assets that are not operating when data is collected results in missed readings, or lower route compliance.
Consider this example: a power-generation facility was facing increasing demand on their peaker plant, meaning that they ramped up during the shoulder hours (morning and evening) when there were no technicians available to collect data. In an effort to ensure asset availability, they installed wireless vibration sensors that enabled them to schedule measurements during the ramp-up, increasing route compliance and data consistency.
The result: within the first six months, they detected 13 developing failures and were able to schedule maintenance to avoid unplanned downtime.
Recent changes make continuous monitoring easier
The last five years have witnessed a step-change in wireless technologies, providing lower costs and greater collaboration, integration and analytics. Four trends have contributed to this:
- Lower sensor costs—With the introduction of MEMS sensors, the cost of wireless sensors has fallen by 75% over the past five years.
- Cloud storage—Native cloud applications have been introduced, providing frequent updates and the ability to access data anytime, anywhere, from any device.
- Machine learning—With multiple sensors collecting readings several times per day, data analysis quickly becomes overwhelming for analysts. Advances in machine learning can serve as an assistant to prioritize assets that need further diagnostics.
- Increased acceptance of wireless infrastructure—As companies invest in mobile workers, they are open to other technologies that can securely leverage existing networks.
Even with all these advances, variable-speed assets have been largely excluded from equipment that can be monitored with wireless condition monitoring. Wireless sensors take more frequent readings, but the measurement interval is scheduled rather than determined by operating conditions, meaning data is collected and transmitted, regardless of whether it is running at the same speed (or even turned on). Sending unnecessary readings can significantly reduce battery life, which is always a key consideration in any IIoT project.
Additional complexities include how varying operating conditions will throw off automated analytics. Machine-learning models are trained to recognize patterns—they need adequate data points (and ability to model) for each speed range to be effective.
The next generation of wireless monitoring for variable speed assets
Recent advances in technology have made wireless monitoring for variable-speed assets possible. Sensors with integrated speed detection enable more efficient data collection and transmission, waking up the sensor only when operating state is on or speed is within a specific range. This means data collection is consistent and trendable, providing cleaner data for analytics and better battery life.
Speed detection opens up additional functionality, such as smart measurements at the edge, in which some computing is performed by the device itself. Consider synchronizing measurements between sensors, where one sensor wakes the others up based on operating state. Or consider optimizing battery life by only taking and transmitting data when the asset is operating at the right conditions.
As companies are increasingly driven by economics and sustainability, they will increasingly switch less efficient drives over to VFD equipment. With the combination of speed, vibration and temperature, continuous monitoring of these semi-critical assets has become possible. This helps to expand IIoT projects to include variable-speed, spared or intermittently operating assets.
And as technology continues to evolve and incorporate more assets, companies can see more use cases for wireless monitoring and increasingly bridge the gap between pilot and full deployment.
Juli Iacuaniello is marketing director with Petasense