KNOWLEDGE ELICITATION and DIGITIZATION USING FRAM to INFORM AUTOMATION of MARINE OPERATIONS in ICE

Autonomous systems are becoming more common in the maritime industries. Some marine operations are more straightforward to automate than others. For example, the navigation of local vessels can be more easily automated on account of well-established standard operating procedures and relatively predictable operating conditions. Complex operations, such as ice management for ocean going ships, are less amenable to automation. Ice management is complex in nature and often takes place under challenging circumstances that include harsh weather, time constraints, and multiple information sources. The conditions are dynamic, and it is difficult to predict precisely how the circumstances will evolve. Successful ice management depends both on prior managerial planning and on-The-job adjustments made by the seafarers in dynamic operational conditions. To reach the level of autonomy equivalent to expert seafarers, the machine needs to assess, reason, decide, and act like them. One way to achieve this is to train the machine by the experts. This paper presents how the Functional Resonance Analysis Method (FRAM) can facilitate such training by connecting the human (expert seafarers) and the machine. A FRAM model of the ice-management operation will be used to 1) understand the ice management operation from experts and visualize it, 2) identify the salient features used and the on-job-Adjustments made by experts during the operation, and 3) guide the digitization of the collected knowledge for icemanagement automation.