The Changing State of the Asset-Intensive Industry
As technologies like AI, Analytics and IoT make inroads into the manufacturing industry, asset-intensive companies are reassessing the best maintenance strategies. While companies are shifting towards predictive asset maintenance from reactive and planned maintenance (more of a scheduled type of maintenance), companies have often limited predictive maintenance to the critical assets due to high costs and inefficiencies in predictive maintenance solutions. With the advent of AI bringing further automation and intelligence, the industry is now reassessing this approach for obvious reasons, such as maximizing uptime, reducing maintenance costs, improving asset performance management and ensuring safety. One of such approaches has been asset criticality- where you rank & prioritize various plant assets that you consider critical to avoid unplanned downtime at the plant.
The Heavy Costs of Downtime
Harnessing technologies like AI and analytics seems a natural choice because unplanned downtime (due to anomalies in machine load, spindle speed, torque, acceleration, etc.) costs time, energy, and most importantly — money (heavy financial losses are attributed to reduced throughput and sometimes even due to the scrapped tools that no longer can be used).
Mitigating risks with Asset Criticality
In a large plant, there can be a large number of issues/alarms at a given time, and it becomes challenging for the plant operators to prioritize repair, maintenance or other corrective action. As the resources deployed are limited, attending to each of the assets can become complex. Furthermore, not every maintenance approach will cost the same. For example, a “run to failure” strategy won’t cost much, whereas “keeping a spare” or using “predictive maintenance software” will involve additional costs.
To identify the right maintenance approach, most asset-intensive organizations generally run a criticality analysis on assets, which is a thorough evaluation to determine which of the assets should be prioritized, and which can wait. At a high-level, it is categorizing assets into two groups: critical and non-critical.
Defining Asset Criticality Analysis for Your Plant
The most common method of defining asset criticality is defining important characteristics for all maintainable assets, and then assigning a weighted score for each of the characteristics for all the assets. The characteristics could be a wide range of attributes like mission and customer impact, the total value of asset replacement, safety, health, environment impact, failure probability, etc.
This assessment should determine the most critical assets, which can help you assign different levels of prioritization (like L1, L2, L3, etc.).
How can AI make a difference here?
Conventional asset criticality management is centered around using sophisticated approaches such as predictive maintenance for the most critical assets. Since high investments are involved (which is proportional to the number of assets being covered), more often than not, enterprises limit predictive maintenance to only a few assets. This means that the risk of downtime caused by auxiliary equipment, excessive maintenance or high spare inventory costs still looms over other less critical assets due to manual and more rudimentary approaches used. Monitoring only critical assets also does not take into account the interdependency of critical systems on the rest of the equipment in the plant. This can often result in downtime regardless of your most critical assets being covered. Furthermore, since only data from critical equipment is available in the predictive maintenance solution, diagnosis the exact issue is also difficult.
To address this shortcoming, an AI-based solution like UptimeAI can help enterprises to bring the entire plant under predictive maintenance. With high levels of automation, intelligence, and self-learning workflows, UptimeAI significantly improves the scalability of the predictive maintenance solution by reducing the monitoring cost per tag/sensor.
Another boon is a 360-degree view of all your interconnected assets. Covering the entire plant makes it easier for AI algorithms and analytics to highlight root causes, diagnose them and offer remedies (which is unlikely when only some assets are covered). This makes sure that your process is treated as an interconnected entity, with all assets monitored constantly.
Conclusion
Asset criticality was a necessity a few years back, when connecting assets and monitoring the full plant with predictive maintenance was costly and inefficient. Manufacturers then had to prioritize the life and performance of a high-value asset over the full-balance of plant optimization due to limitations in the solutions. With an effective AI solution in the picture, you don’t have to just protect critical assets and hope the best for the rest. You can now predict problems and understand causes for the entire fleet. The result is a plant (not limited assets) that is reliable and efficient with minimal risk and maintenance costs.
All in all, 2021 promises to be an interesting year for Manufacturing, as the industry evolves into one that is innovative, agile & responsive both to the needs of consumers and the market, that will take it to newer heights.