An AI For Ship & Yacht Design: Design, Design Engineering, Company ERP, Decision Maker (Although it's relatively slow innovating industry)

An AI For Ship & Yacht Design: Design, Design Engineering, Company ERP, Decision Maker (Although it's relatively slow innovating industry)

-    AI tools are not approaching, it's already here and it's being widely adopted in all industry, as always of course ship and yacht building.

-    And it is not fair to blame this industry!! It can be explained why this market is very small compared to others, but the main argument here is, it's quite logical that this industry hasn't even started to adapt AI tools.

To understand the scale, consider the below text:

"The shipbuilding and yacht building industry, with a market size of around $287.14 billion in 2023, is relatively small compared to other top global industries. For example, industries like software (system & application), semiconductors, and online retail have significantly higher market caps and revenues. The software (system & application) industry alone has a market cap of $6,613 billion and revenues of $611 billion, while the semiconductor industry has a market cap of $4,259 billion and revenues of $664 billion. Online retail boasts a market cap of $2,812 billion with $740 billion in revenue. These figures starkly contrast with the shipbuilding and yacht building industry's market size, underscoring its relatively smaller scale in the global economic landscape​​."

Now let's dive into the proposal.

• Step 0: Find a Great Large Language Model (LLM) - OPEN SOURCE 

• Step 1: Fine-tune this LLM based on a shipyard data and notes that are within the context.

• Collect questions and answers
• Collect entire emails exchanged in that company, phrase it to become questions and answers.
• Gather all the IACS classification rule books. Add it to the dataset.
• Gather all the books on ship, yacht design, shipyard management, design engineering principles.
• Make sure your data is around 10GB in size, for large scale fine-tuning.

• Step 2: Start the Training!!
• Make sure you have a solid hardware.

• Step 3: Start Evaluating the Model
• Run Tests

• Step 4: Build a Generative Design AI
• Gather designs of yachts (in this case) in 3D
• Gather designs of construction in 3D
• Gather designs of workshops for that particular project
• Do this for other parameters.
• Enlarge the dataset.
• Train it!!

• Step 5: Communicate this AI model with the first model that you created
• And there you have an AI system that can design engineer a premature ship or yacht based on the IACS rules

• Step 6: Improve it by Reinforcement Learning and Supervised Learning Methods such as from Human Feedback