From 250 ideas to 27 AI use cases at Nidec Drives.

Delivered by Enceladus AI, the PureNet Group’s AI consultancy.

The challenge

After a leadership briefing had given its senior team a common understanding of AI, Nidec Drives wanted to go further: find where AI could help across the whole business, and choose where to start.

The briefing had changed the question to where AI would create the most value, and how to approach each opportunity. Answering that for a business as broad as Nidec meant hearing from the people doing the work in every function: where it hurts, what AI might do about it, and how ready the business really is.

The answer also had to drop straight into Nidec’s own technology-investment process, so the work ended in decisions, not another report on a shelf.

The project

This was the discovery and prioritisation stage of Enceladus AI’s Converge Methodology, run in full across five areas: R&D, manufacturing, procurement and supply chain, sales and service, and marketing, with product management and IT taking part in each.

Five workshops to open things up. One per function, each gathered every idea for where AI could help, without judging feasibility yet. Each idea was recorded with the problem or opportunity behind it, how the work is done now, the kind of AI it would use and who would gain. They ranged from AI built into the drives themselves, through engineering tools such as AI-assisted coding and design, to the everyday knowledge work every department shares.

Five workshops to narrow them down. Before each, Enceladus AI grouped the function’s ideas and gave every group a first score against ten readiness measures, on purpose a starting point for the room to argue with and rank. The final scores were the business’s own.

The same needs kept appearing in different functions: an assistant to draft and summarise, a way to search knowledge spread across systems, the wish to forecast demand or watch the market. These became cross-business themes, shared with every group.

What was delivered

Ideas that relied on the same capability were combined into single use cases ready to submit: eleven cross-business themes and sixteen candidates for single functions. Each was written in Nidec’s own assessment format, with the problem, the opportunity, what it brings together, its scores, its route to adoption, its place on the AI Opportunity Pyramid, and possible suppliers drawn from more than a hundred reviewed across twelve functions.

The roadmap sorted them by readiness (ready now, nearly ready, strategic, and research) with the data foundations running beneath all four, and a recommended first quarter followed from it. The full record of all 250 ideas, with how each was captured, grouped and scored, was handed over as a workbook, so every decision can be traced back to the workshop it came from.

What we found

Data is the limit, not technology. Tools already exist for most of the opportunities. What decides whether one can be delivered, and how soon, is whether its data exists in a usable form. Only about a third of the ideas had their data ready; the rest needed research or data groundwork first.

Enthusiasm and readiness pull apart. The themes with the most ideas were often the least ready. So the roadmap follows readiness, not excitement, and starts the master-data and knowledge foundations early, because so much rests on them.

The biggest blocker was the cheapest to fix. Restricted access to AI tools was the loudest complaint in every workshop, and it stood between the business and its quickest, most valuable wins. The worry behind it was fair; the fix is governed, approved access rather than a blanket ban, which only drives people to unapproved personal tools.

The programme turned a wide field of ideas into something the business could act on: a prioritised set of use cases, a clear sequence for delivering them, and a first quarter that shows visible value while the foundations for the larger prizes are laid.

What’s next

Nidec’s participants use the report to confirm the use cases and their order, and the business owners take them into the investment process. The recommended first quarter: approve governed access to AI tools, roll out the two quick switch-on wins, start proofs of value where the data is already in hand, begin the data foundations alongside, and adopt the responsible-AI framework. Piloting and scaling, the next phase of the Converge Methodology, follows for the opportunities chosen to go ahead.