ENABLING CONDITION MONITORING WITH PREDICTIVE ANALYTICS

Transcription

1 ENABLING CONDITION MONITORING WITH PREDICTIVE ANALYTICS Predictive analytics can foster real-time machinery condition monitoring for asset-intensive manufacturers, paving the way for increased operational efficiency, enhanced resource utilization, and robust risk management. ASHOK KUMAR Head - AI & Analytics

2 Table of Contents An Industry Perspective 03 Predictive analytics: A superior method of equipment maintenance 05 Challenges 06 Reliability-based monitoring 08 Conclusion 09 Reference 09 About L&T Technology Services 10

3 AN INDUSTRY PERSPECTIVE Diagram 1 Dawn of Industry 4.0 The manufacturing industry is entering a new era as an unprecedented confluence of disruptive digital technologies heralds the dawn of Industry 4.0, where the physical and digital worlds converge to transform the entire value chain. On the other hand, political uncertainties, sluggish economic expansion, ever evolving regulations, and anti-globalization movements worldwide are posing dynamic challenges for the industry. In such a scenario, reducing operational expenditure (OPEX), superior cost efficiency and resource utilization are increasingly becoming the top drivers for competitiveness 1. While many manufacturers were earlier able to reduce their overheads and keep up production through outsourcing to lower-cost countries, the same approach may not prove to be optimal going forward. As policy makers in several developed countries and even some developing ones respond to growing nationalist sentiments with regard to trade and commerce, industrial equipment makers today face inward looking policies and increased import duties. This reality makes it imperative for manufacturers to revisit their core production principles. There is now an urgent need to evaluate different ways of decreasing operating costs and improving asset productivity. 1 PAGE 3

4 WHY EQUIPMENT MONITORING MATTERS Typically, manufacturers allocate 40% of their operating expenditure toward running critical equipment, with an additional 5-8% earmarked for maintenance of the same. Any unexpected equipment failure not only increases the cost of operations but also affects asset utilization levels adversely. Plant equipment monitoring thus assumes significant importance for asset-intensive manufacturers to minimize the uncertainty around maintenance and reliability. THE GAPS IN TYPICAL MAINTENANCE APPROACHES TODAY Efficient and fail-proof monitoring and maintenance practices go a long way in improving factory floor operations, and reducing unforeseen expenditures. Currently, manufacturers primarily employ two types of maintenance corrective and preventive. In the first method of maintenance, machines are repaired only when they break down and/or are not performing up to the mark. This is not a very effective approach as it can cause substantial disruption in scheduling and production owing to the delays in sourcing spare parts, significant time taken to repair, and getting access to machinery experts. There is also a high risk of damaged machinery causing accidents and injury to workmen, and failing safety standards. In preventive maintenance, equipment is attended to on a fixed and periodic basis, depending on predetermined breakdown windows. The frequency of such monitoring is decided based on the estimated rate of equipment wear and tear, acceptable maintenance costs, and general degradation rules. Most manufacturers today rely on preventive maintenance to ensure smooth operations. However, machines can be unpredictable, and even break down right after a scheduled maintenance. Over the years, there have been many such cases of unanticipated collapses that brought down entire factory floors to a halt. Such risks, if not addressed proactively, can cause extensive business damage, both in financial and reputational terms, apart from endangering the safety of workers. With various digital technologies transforming the manufacturing business in a fundamental manner, continuous production and sustained profitable growth has become critical. Manufacturers, therefore, must look beyond preventive and corrective maintenance. There is a pressing need for them to embrace a next-generation, predictive equipment monitoring and maintenance strategy. Through this, they will be able to gain increased control over their production schedules and minimize operational uncertainty to a great extent. PAGE 4

5 PREDICTIVE ANALYTICS: A SUPERIOR METHOD OF EQUIPMENT MAINTENANCE As the Internet of Things (IoT) increasingly becomes mainstream across the industrial landscape, manufacturers have aggressively stepped up installation of machines equipped with sensors throughout their plants. Against this backdrop, some companies are exploring predictive maintenance, or conditionbased monitoring and maintenance (CBM). CBM is an emerging method of equipment care that analyzes the real-time data collected from machines to forecast potential breakdowns. Under the predictive approach, sensor-embedded machines continuously capture equipment data concerning various performance and environmental parameters, and transmit the same to a central processor. The processor uses an advanced analytics engine to detect variations over time and flag anomalies. Factory personnel can then use these alerts to schedule maintenance before a failure actually occurs, but only after a potential cause is identified. This way, while machinery performance is restored ahead of a breakdown, unnecessary time and efforts are no longer spent in tending to equipment that do not require any intervention. Moreover, manufacturers can harness the predictive data mining setup to undertake root cause analysis around various factors that can possibly trigger failures. This way, they will also be able to prevent occurrence of similar defects in the future. Diagram 2 - A superior method of equipment maintenance With CBM, manufacturers can detect outages early enough to fit into their production schedules as maintenance is based on the actual machine condition and not on a calendar. Maintenance planners can also use the extensive information available from the CBM system to stock the right spare parts. They can avoid incurring expenses on storing unnecessary items, and source components that are frequently needed. By delivering all these benefits, preventive maintenance, on an aggregate basis, facilitates better asset utilization and lower total cost of ownership (TCO). By enabling robust risk management, it will foster business continuity and enhanced operational efficiency as well. PAGE 5

6 CHALLENGES On the flip side, CBM can be fairly expensive to implement, especially where the cost of replacing equipment itself might be lower than fitting them with monitoring devices. However, if manufacturers take this skewed view and roll out CBM for expensive equipment only, they might not accrue the full range of benefits associated with preventive maintenance, on account of insufficient machine coverage. Manufacturers with aged machines might also find it unviable to retro-fit old equipment with modern sensors. Additionally, outdated hardware can also lead to inefficient functioning of the overall CBM system. Even if industrial enterprises were to overcome such challenges, they would still have to effectively train their personnel on the new maintenance system, and ensure the data being captured is used as intended. They will also need to make sure the collected data is not isolated from other in-house IT systems requiring additional manual efforts to come up with appropriate outage windows. In spite of these issues, CBM still stands out as a superior way of driving efficient equipment maintenance, since it can boost the plant production capacity while driving down operating costs over the long term. However, it is essential for manufacturers to implement CBM correctly to achieve economies of scale. TAKING A HOLISTIC VIEW With innovative semiconductor manufacturing and digitization-enabled technological breakthroughs helping bring down the cost of sensors significantly, cost-effective CBM solutions that bundle both hardware and software are now a reality. These sensors can monitor a wide range of functional equipment parameters including temperature, acoustics, pressure, vibration and load in real time, and generate an accurate and highly reliable predictive maintenance forecasting model. Companies can address some of the other challenges involved in implementing CBM by taking a holistic view toward equipment health monitoring. Specifically, manufacturers should consider adopting reliability-based monitoring for real-time, predictive maintenance. PAGE 6

7 Diagram 3 - Wide range of sensors to monitor Equipment In this advanced method, machine data sourced from the CBM system is integrated with visual inspection data to build a superior and granular data set for analysis by the predictive engine. The comprehensive data set is then automatically transferred to the Cloud and enterprise resource planning (ERP) applications. The data is further enriched by overlaying production schedules and preventive maintenance schedules from ERP systems. Historical equipment performance data, wherever available, is also factored into the final analysis alongside the real-time data, to uncover patterns and behavior based on a fault model library and machine learning algorithms. Such an approach ensures the results churned out by the predictive analytics engine are highly precise, accurate and real-time. In order to help convert the analysis into action on the ground, the CBM system also needs to have in-built alarm mechanisms that can alert the maintenance personnel to take corrective actions before equipment breakdowns. Faults should be categorized into various severity buckets with pre-defined turnaround times for alerting relevant stakeholders right from maintenance staff and supply chain departments to spare-part vendors and warranty departments. The issue of training personnel in different locations can be tackled by delivering the insights in the form of simple charts and dashboards. The analysis can be aggregated for subject matter experts to access and study remotely. PAGE 7

8 RELIABILITY-BASED MONITORING DRIVING PROACTIVE DECISION MAKING And, the business benefits manufacturers can reap by installing CBM systems designed around the concept of reliability-based monitoring are significant and tangible. To begin with, they can leverage such a setup to detect equipment failures early on, minimize total downtime, and boost overall asset performance. By optimizing maintenance schedules into single shutdowns, companies can also execute downtimes and production schedules in a more efficient manner. Overall, reliability-based monitoring can help manufacturers automate the entire orchestration of events from sensing performance data, updating critical systems and predicting breakdowns to alerting appropriate personnel and triggering schedule repairs with regard to equipment maintenance. This makes advanced CBM a compelling proposition for modern manufacturers seeking to improve asset productivity and cost competitiveness. According to various industry benchmarks, CBM setups can help slash maintenance costs by up to 12% in the first year, and drastically improve machinery availability by as much as 92%. They also typically reduce unexpected failures by about 25%, with repair and overhaul time almost halved. The need for stocking spare parts inventory can go down by 20%. Apart from these quantifiable benefits in relation to asset performance, CBM delivers multiple intangible benefits. It fosters smooth running of factory operations and optimized production, by minimizing plant interruptions on account of machinery-related delays. Some of the other qualitative gains companies can derive include higher customer satisfaction, superior capacity management and better supply chain relationships. PAGE 8

9 CONCLUSION Ultimately, predictive maintenance can help manufacturers become far more proactive and take control of their production to boost operational efficiency through increased resource utilization, and manage risks better. By empowering factory personnel to plan work more accurately, CBM also indirectly helps lift employee morale and performance. Put simply, if designed and rolled out smartly, predictive analytics-driven equipment monitoring can enable manufacturers to improve asset reliability, reduce downtimes, contain costs, and increase profitability. PAGE 9

10 REFERENCES Global Manufacturing Competitiveness Index, Deloitte For more information visit us at Reach us at About L&T Technology Services L&T Technology Services Limited is a subsidiary of Larsen & Toubro Limited with a focus in the engineering services space, partnering with over 50 Fortune 500 companies. A leading pure-play Engineering, Research and Development services company, we offer design and development solutions through the entire product development chain, across various industries such as Industrial Products, Medical Devices, Transportation, Telecom & Hi-tech, and the Process Industry. We also offer solutions in the areas of Mechanical Engineering Services, Embedded Systems & Applications, Engineering Process Services, Product Lifecycle Management, Engineering Analytics, Power Electronics, Machine-to-Machine (M2M), and the Internet-of-Things (IoT). Copyright 2018 L&T Technology Services