The Maintenance Impact.
Maintenance costs constitute a major part of the total operating costs of manufacturing plants but very often these costs do not bring the desired effect. Investigations on maintenance management effectiveness state that 33 cents out of every dollar spent for maintenance are spent in vain due to utilization of inappropriate maintenance programs. Such ineffective approaches result in substantial loss of profitability. Furthermore, these poor maintenance programs are the reason for which many plants fail to produce high-quality products, and therefore lose their ability to successfully compete on the market. The dominant reason for this ineffective management is the lack of factual data to quantify the actual need for repair or maintenance of plant machinery, equipment, and systems.
Predictive Maintenance philosophies today – is it really predictive yet?
Traditional philosophies such as run-to-failure and preventive maintenance are now being replaced by predictive maintenance and proactive philosophies. These new approaches are developed from the traditional, so it is necessary to know their strengths and weaknesses.
Breakdown or run to failure maintenance: based on a reactive logic, no efforts are undertaken to anticipate maintenance requirements - machinery is allowed to run to failure and is only repaired just before or when the equipment comes to a complete stop. This is the most expensive approach associated with high expenditures for overtime labor, spare parts inventory, and with losses due to high machine downtime, and low production availability.
Preventive or time-based maintenance: based on a time-driven approach, maintenance activities are scheduled at predetermined time intervals, based on calendar days or runtime hours of machines which usually results in performing maintenance tasks too early or too late. As a result, there is a high probability of unnecessary repairs or catastrophic failures.
Predictive or condition-based maintenance: based on a condition-driven logic, the basic idea is to determine the optimal time intervals between repairs and the minimal number of unscheduled outages. Since the majority of mechanical problems can be substantially mitigated at early stages, predictive maintenance management aims to identify problems in advance, before they actually become serious.
Proactive or prevention maintenance: based on root cause failure analysis, it combines all of the predictive/preventive maintenance techniques with root cause failure analysis. This approach identifies and pinpoints the influencing factors that cause defects and aims to establish proactive actions for avoiding recurrence of such problems.
Both predictive and proactive maintenance philosophies are data-driven approaches. Fitted with the appropriate sensors to measure vibrations, pressures, speed changes, temperatures, torques etc., production machines create endless streams of data and provide an accurate representation of the operating condition of a single machine or system. To predict the operating health conditions, manufacturers need to systematically process and catalogue the data since predictive models require historical records to make assumptions about a machines’ future. Unfortunately, many manufacturers seem to have misplaced their ticket on the big data train – and without appropriate data sources in place, predictive modeling is not an advisable way to go – and presumably not established yet as key technology within condition monitoring programs.
Predictive Maintenance is data driven.
For us, data driven means that progress in an activity is compelled by data, rather than by intuition or personal experience. Most of today’s condition monitoring measurements are executed manually on the shopfloor by engineers using portable devices – with functionalities to export both a snapshot of raw data and corresponding condition measures. But data streams beyond this measurement time window remain hidden and unknown. With big data technology in place, the real time and high-frequency nature of sensor and machine data can now be processed and analyzed in a very efficient way – 24/7 – fully automated – in real-time. Keeping in mind that fatigue-induced-cracking often happens without prior warnings, the benefit of a nonstop data driven acoustic emission monitoring is obvious. Changing states of machine conditions can be monitored in real-time, using anomaly detection - future machine failures can be forecasted with prediction models in place. Nowcasting, the ability to estimate metrics such as root cause analysis immediately, something which previously could only be done retrospectively, is ready to becoming more extensively used, adding significant power to prediction. A permanent frequency of data allows users to test theories in near real-time and to a level never before possible.
Sensor orchestration - a one stop shop for condition monitoring. DATATRONiQ utilizes bleeding edge big data technologies and machine learning algorithms and puts them to work for factories, solving perpetual challenges in efficiency and quality. It is provided with a unique plug & play connector we typically connect directly to the machine controller permitting access to all of the machine’s data. Additional vibration sensors can be easily fitted to the machines – even antiquated production machines can be part of the DATATRONiQ data universe within minutes by retrofitting with our vibration sensors. Our technology makes it possible to affordably collect and process huge volumes of data from hundreds of sensors and machine controls in a single production plant, analyze that data in real-time and detect problems before they actually happen. The combination of anomaly detection, root cause analysis and failure prediction creates actionable insights and provides a complete picture of all machine conditions and condition monitoring activities. Prebuilt condition monitoring metrics are unified in dashboard views, can be transferred to shopfloor (e.g. MES) or topfloor applications (e.g. ERP) or directly interfere with machine controller in real-time to shut down a machine in a critical state.
Quick wins and benefits of data-driven condition monitoring.
There are many reasons why manufacturers should establish a holistic maintenance program to monitor actual operating conditions of the plant equipment and systems to optimize the total plant operation. Benefits range from increase in machine productivity, extended intervals between overhauls, improved repair times, and increase in machine life up to improved product quality. Suffering from enormous efforts to manually gather, measure and analyze the relevant data to assess a machine’s condition in the past, big data technology now makes data streams accessible, processable and actionable in a very efficient way. With DATATRONiQ, organizations can significantly reduce their maintenance costs by applying data-driven maintenance strategies to innovate, compete, and capture value. Even small enterprises can now join the big data train – and listen to the heartbeat of their machines in a very affordable and proficient way.