Modern defence platforms, such as tanks, corvettes, aircraft carriers, submarines, or aircraft, function thanks to a multitude of software, hardware such as sensors and actuators. These systems are becoming increasingly complex, interconnected and interdependent. At the end of the day, they form a huge system of systems, characterized by tens of thousands of real-time data sources and data channels. The amount of data they produce is massive. The key characteristic of the said systems is their immense complexity.
Because of technology advances the complexity of defence systems is growing quickly and becoming a threat in itself. This is because high complexity implies fragility. Highly complex systems can suddenly behave in an unexpected manner. As complexity increases this becomes more likely. The QCM has been engineered specifically to monitor in real time the complexity of mission-critical systems and infrastructures and to identify within the said systems any concentrations of fragility which can impair their functionality.
But QCM has also another function. As it monitors systems or infrastructures it identifies sudden peaks in their complexity which tend to precede anomalies or malfunctions. Such peaks constitute the core of QCM’s anomaly and crisis anticipation capability. The beauty of the approach lies in the fact that it is not necessary to dispose of hundreds of examples of an anomaly in order to be able to detect it. This would be very expensive and impractical. Besides, as complex defence systems evolve, so do their anomalies. Keeping track of all of them is unimaginable. In essence, QCM doesn’t need to learn. All it does is detect jumps in complexity.
Because of the high operational and maintenance cost of defence systems, it is paramount to constantly monitor them in order to prevent malfunctions and breakdowns, as well as to guarantee mission readiness. In virtue of the immense complexity of these systems, maintenance schedules and interventions based on mileage or engine hours, is too simplistic. However, monitoring a system such as a warship, by observing and recording the values of each sensor and making sure that they fall within nominal bounds is not enough. QCM offers a more advanced Condition-Based Maintenance mechanism in that it takes into account real time systemic and holistic aspects by analysing interdependencies between all subsystems and data channels. This takes system monitoring to a whole new level. Health monitoring of complex systems must take into account their key systemic characteristic – their immense complexity.
Condition-Based Maintenance (CBM) is a maintenance strategy that monitors the real-time condition of an asset in order to determine what maintenance needs to be performed. Unlike preventive maintenance, which uses things like calendar-based maintenance or other means to determine when to schedule and perform maintenance, condition-based maintenance dictates that maintenance should only be done when these real-time indicators show irregularities or signs of decreasing performance.
The goal of condition-based maintenance is to continuously monitor assets to spot impending failure or anomalies, so that maintenance can be proactively scheduled before the failure occurs. The idea is that this real-time monitoring will give maintenance teams enough lead time before a failure occurs or performance drops below an optimal level.
At a very high system-level representation, QCM monitoring accomplishes a mapping of raw streaming data onto the so-called Complexity Map, illustrating the instantaneous interdependencies between the various subsystems and their relative footprint on the State of Health of the entire system. In the map, large boxes represent subsystems that, at a given time, are the drivers of the state of health of the whole system and are being “stressed” more.
Complexity Maps are paramount when it comes to forensics, i.e., finding the sources of failure or malfunctions after a breakdown.
There are many contexts in which an anomaly may be fatal. Many of our clients don’t have the luxury of being able to survive hundreds of anomalies only to be able to train a system to recognize the next one. Conventional anomaly detection utilizes Machine Learning to teach a system to recognize anomalous situations. However, in highly complex systems there exists thousands of possible modes of functioning and for each mode there may be as many potential anomalies. As complexity increases, these numbers increase too. So, how does one detect an anomaly in such contexts? What is an anomaly in such circumstances? Clearly, the classical approach won’t work.
QCM suggests a different solution. We know that rapid complexity fluctuations (spikes) anticipate (or accompany) phase changes or mode transitions in dynamical systems (problems, in plain English). In many cases they provide a formidable early warning signal. An example of complexity evolution of a complex system is illustrated in the figure below. Peaks correspond to anomalies that QCM has identified. Once an anomaly is found, the system indicates which sensors (data channels) are its drivers/causes. This does not require any form of Machine Learning.
This is performed by the Complexity Profile which ranks components according to their contribution to system complexity. This allows to determine which of the channels are responsible for the increase in complexity, as shown in the example below.
The Anomaly Early-Warning Dashboard is illustrated below.
0 comments on “Condition-Based Maintenance”