Adapted from the update iNDustry Labs sends the AI Executive Group, a group of manufacturing and logistics leaders across the South Bend–Elkhart region.
What companies actually asked for
Seven AI projects launched this cycle. The problem statements below are the companies' own, lightly edited, and they are a good read for anyone wondering what a realistic first AI project looks like. Every one of them starts from a specific operational irritation rather than from the technology.
Querying the data warehouse in plain language
"When we get unique customer inquiries such as how many parts do you make that are under three pounds, we spend a lot of time writing complex queries, often by people who are not SQL-trained. If a chatbot could query our data warehouse from natural language input, it would save us a lot of time and give customers faster answers."
A multilingual HR assistant
"We have many employees who struggle to interpret large HR documents, policies and handbooks, and many who do not speak fluent English. If we could build a multilingual chatbot that uses those HR documents as its backbone, employees could get answers directly, which is a better experience for them and saves our HR director a lot of time."
A financial planning and analysis agent
"Our accounting team has limited time to analyze financial documents and compare them to forecasts. Implementing a full FP&A system is costly and time-consuming for a medium-sized business like ours. We want to explore whether AI agents can deliver the same value while making documents more accessible through natural-language summaries, and to test the agent's accuracy and deployment speed against commercial FP&A tools."
Forecasting automotive accessories
"We have begun exploring how machine learning models can improve forecasting for the automotive accessories we sell. We have had some success with models such as XGBoost for certain product families, but we struggle to forecast others. Finding out how much better we can get would resolve one of our largest organizational challenges."
Automating the triage of new sales leads
"We receive most inbound sales leads through a web form that requires manual review and entry into our CRM, with a sales team member verifying each contact's legitimacy and suitability. An agent that analyzes submissions, does brief research and prioritizes leads for review would save significant time. We would also like qualified lead details populated automatically before a person gets involved."
A project management agent
"Hot orders and customer response delays force significant re-prioritization of our project plans and individual staff tasks, which often puts tasks on the critical path. If we could use agents to manage that without adding project-management headcount, we could improve delivery."
Automating apprentice performance tracking
"Apprentice performance, field touchpoints and retention outcomes are spread across several systems and spreadsheets. Supervisors do not have a consistent single view of progress, and they spend significant time manually compiling monthly coaching updates and program reporting. The goal is to consolidate that data and generate supervisor-ready monthly updates."
Every one of these started from a specific operational irritation, not from the technology.The through line across all seven scopes
From the events this cycle
The Q1 AI keynotes at Innovation Park paired a session on measuring AI return on investment, from Notre Dame's Mendoza College of Business, with a regional company walking through its own AI journey, followed by a student product design prototype showcase. At IDEA Week, a panel on AI in manufacturing brought together executives from two regional manufacturers.
An earlier session on attracting AI-savvy talent drew HR leaders from across the region, with speakers from a regional manufacturer and from Notre Dame's HR and technology groups sharing what they have learned about hiring for AI fluency.
Upskilling without a STEM background
Notre Dame's Lucy Institute launched ExLENT Data Crossings, an initiative to upskill students and working professionals in data science. It is aimed squarely at employees who lack prior STEM education but show strong potential with data, which makes it a useful route for companies trying to build capability from the team they already have.
Shared by group members
From the ProveIT 2026 conference, a member shared an AI-enabled smart factory dashboard from a real factory running inference on live data streams. Worth a look if you are trying to picture what continuous inference on plant data actually produces.
iNDustry Labs runs the AI Executive Group alongside the AI Leaders of Tomorrow fellowship at the University of Notre Dame.