FAULT PREDICTION AND PROACTIVE MAINTENANCE

Nowadays, naval engineers and shipping companies try to constantly minimise fixed and operational costs, as well as their impact on the maritime and generic environment. Amongst the main reasons that tend to significantly increase costs are unpredicted damages and/or mechanical failures. By providing a simple overview of the effect of such phenomena, one can report high costs of repairs and spare parts (especially when requested in as close to real-time as possible), loss of earnings due to immobilisation of vessels and/or due to failure to comply with Service Level Agreements (SLAs), or even port state control (PSC) detention. Indirect costs can be foreseen as well; loss of reputation being amongst the most important. The same (unwanted) phenomena can lead to environmental pollution, as leakages and increased emissions are a significant result of unpredicted damages and/or mechanical failures (although modern technologies have limited the chances of these types of side effects to a large extent).

Fault prediction and proactive maintenance

Along these lines, in order to minimise such phenomena, contemporary ships and vessels in general are equipped with a plethora of sensors and monitoring utilities, constantly collecting operational and performance data on every critical aspect of the ship’s operation (e.g. engines’ strain, emissions, fuel consumption, load), regardless of the nature of the ship (e.g. cargo ships, transport ships). However, although sensors are present, their full potential is not frequently exploited. Not all incoming data is stored, mainly due to space (memory) shortage, and maybe most importantly the sensors’ data is not integrated or correlated with historical or external data streams.

Challenges: The previously described process has undeniably benefited the timely identification and prediction of upcoming damages or mechanical failures. However, one core aspect remains unsolved: processes and systems like the previously described are vessel-centric. All models, calculations and estimations are based solely on data coming from the vessel itself. Even if the utilized systems support real-time data collection, the process needs to take into account past events and entries in order to provide an early warning or verdict.

The question that remains unanswered is whether there is a way to embed additional relevant parameters into the process, making the proactive confrontation of damages or mechanical failures more effective and accurate; and thus providing important financial and environmental benefits.

Expected Benefits: The first set of benefits fully corresponds to the goals of the initiative. The company envisions minimizing repairs and maintenance, as well as minimizing loss of earnings due to the vessels’ inability to operate, through this initiative. In addition, environmental impact is also envisioned to be reduced, since fewer damages or mechanical failures during operation will result in practically eradicating the occurence of unpredicted spills, emissions etc. Moreover, indirect benefits could include commercialising the enriched data and events’ portfolio and/or prediction model, targeting other industries in the sector, as well as building a reputation around reliability and innovation.

Comments are closed