AI in Maritime Shipping

Few weeks back, I did a rapid survey/reading on AI applications for Maritime shipping industry to teach the students of the online MBA (Digital Maritime & Supply Chain, #IITMadras). The requestor drew the boundaries on the topics and guidelines, by scoping - AI in maritime, low-code-no-code (LC-NC), #RAG/LLM, and entrepreneurial Biz—opportunities, interactive-lab-activity, and of course all within the stipulated time. I added low-math-no-math as another condition.

My search led me to a subset of perennial maritime challenges and suitable AI solutions which I could share with the students. Some of them are Reinforcement Learning (#RL) based optimal ship routing, fuel efficiency, cargo theft, CV, Maritime Data& analytics, and weather forecasting to name a few. I narrowed down a few topics because shipping giants like Maersk and others have been benefitting from such AI solutions. 

I could show them LC/NC through products – H2O.ai (Open), Data Robot (DL), Robo Flow (CV), Rapid Miner (data wrangling, student license), Azure #databricks, and #DataFactory. Replicating an existing RAG solution (Raja Vasanthan, a professional Tamil video) myself in our lab (hack project) gave me the clarity to explain in simple terms the working of RAG/LLM solution to the Biz. students. This was a true co-learning experience. Narasimman helped responding to questions-stream. Magesh had setup the Azure Big Data. Raghavendra Prasad, John Robert, Ramesh Singhal & Kavita moderated.