Increasing Uptime and Performance by Using Digital Twins in Dredging Diagnostics
Type:
Presented during:
WODCON XXII - Enhance the Harmony between Dredging and Ecology
Authors:
J. Osnabrugge and B. van den Berg
Abstract
"Currently available diagnostic information for dredging equipment and vessels is often basic and insufficient for fast troubleshooting and problem solving of failures. As onboard automation systems are getting more intelligent and integrated they also become more complex. For this reason it is important that diagnostic capabilities regarding dredging vessels and equipment also improve in line with this overall system complexity. In case a component such as a sensor, actuator or automation controller is not working correctly, it may have significant impact on the performance and uptime of the total dredging equipment and vessel. Hence, it is important to identify and prevent failures in an early stage and make this information available for fast troubleshooting and preventive maintenance actions to minimize or prevent loss of performance and uptime. This paper presents a, for the dredging industry, new approach to developing a diagnostic system for dredging vessels and equipment. The basic concept is that parallel to the real dredging vessel a simulation reference model containing the expected ‘behaviour’ of the dredging vessel, named the ‘digital dredge twin’, runs in the diagnostics software. This digital dredge twin is built based on a combination of process knowledge and physics, data analytics and artificial intelligence algorithms obtained with the development of training simulators and advanced automation systems in the last decade. When the measured ‘behaviour’ of the real dredging vessel deviates from the digital dredge twin the diagnostic system is triggered and analyses the cause of this deviation. This may be, for example, actuator or sensor offsets, slower actuator dynamics or loss of efficiency. Several diagnostic examples are presented in this paper like dredge pump performance diagnostics and sensor diagnostics of the mixture transport process. An additional feature of the diagnostic system is the use of virtual sensors. In case of a sensor failure a virtual sensor from the digital dredge twin can be used and temporarily replace the failed sensor to prevent loss of dredging uptime or performance. Real-life usage examples of this virtual sensor feature are shown for mixture density and mixture velocity sensors."
Keywords: artificial intelligence, dredge pump performance, virtual sensor, advanced automation. INTRODUCTION