Eye on Innovation: Shipyard AI Contest Part of Growing AI Trend on the Waterfront
- Category
- Military affairs
- Source
- Norfolk Naval Shipyard
Norfolk Naval Shipyard:
"Production Workforce Development Director (Code 900B) John Rowe recently initiated a challenge to the production workforce. He asked, “Who can come up with the best use of AI to make the job easier and positively impact the mission?” The internal innovation challenge was to find practical, everyday AI solutions to streamline shipyard workflows and support the Navy’s mission.
The winners, John Karsten, Monique Reese, and Stephen Campbell, were featured in NNSY’s “Talking Ship” video series in May on the shipyard’s Instagram page.
First-place winner John Karsten leveraged AI tools to quickly generate contractual Statements of Work (SOW). “I use AI to write a lot of technical documents and Statements of Work, and Independent Government Estimates (IGE),” said Karsten. “I can tell you right now it saves me about 95 percent of my work hours that I used to have to spend expending writing these documents that I can just plug into AI and AI generates what I need. So, if you’re not using AI in the shipyard, you’re definitely behind the power curve.” The use of AI as a technical writing assistant has increased the acceptance rate of SOWs and IGEs to 90 to 95 percent, compared with 0 percent before using AI.
Second-place Monique Reese developed a Standard Operating Procedure (SOP) agent and reference workbook. “I’ve created a standard operating procedure agent with AI that will help us reduce creating and revising SOPs from weeks to hours,” said Reese. “Not only that, but it will also align a workbook that will step-by-step, play-by-play the entire process for waterfront workers.” This is useful for workshops that haven’t formalized their best practices in an SOP. The agent automates and simplifies the workflow, allowing users to submit inputs and generate an SOP almost instantaneously. Before the use of AI, the time and resources required were substantial and deprioritized.
The third-place winner was Stephen Campbell, who employed AI to automate and streamline Team Learning Session (TLS) analysis and its required documentation. “I create a three-agent AI team for TLSes. This allows us to do timeline establishment prior to a TLS and saves us days allowing us to get to the heart of the matter,” said Campbell. The intent is to have agents act as impartial AI investigators to perform root cause analysis for unplanned events applying Lean Six Sigma principles.
Coordinating the AI contest was Code 960B Chris Gallihugh. A total of 24 submissions were received for the best uses of AI agents. The ideas ranged from using AI for solving simple math problems quickly to attempting to revolutionize the way technical work documents are generated. Each presented its own special set of challenges regarding implementation.
“Our work often demands reviewing extensive written procedures and complex technical references, a process that traditionally took days of manual searching across shipyard instructions, NAVSEA guidance, OPM policies, and federal regulations. With AI, that multi-day comparison and research process now takes just a few hours,” said Gallihugh. He also cautioned, “There are strict prohibitions with the types of material that can be loaded into Gen AI. Users should understand those prohibitions before using the tool.”
The Production AI contest is part of a larger, ongoing command push to introduce artificial intelligence, machine learning and advanced digital workflows directly to the waterfront. Other recent efforts include Gen AI command climate analysis, Amelia & IT intake, and the Department of War employment of AI large language models (LLMs) for personnel to use. In August, the Shipyard Resilience Team developed a Defense Organizational Climate Survey (DEOCS) desk guide with the assistance of Gen AI to provide timelier analysis of results so that leadership can act on those improvement areas. In June, the NNSY introduced Amelia, the Navy’s virtual AI agent to provide 24 hours a day, seven days a week tier-0 self-service and tier-1 initial agent support for IT assistance. The Navy has three AI platforms available for shipyard employees including ChatGPT, Gemini, and Grok. These platforms assist with streamlining workflows, generating ideas, and supporting the Navy’s mission."
Source: Norfolk Naval Shipyard