Where Does AI Work Best With Lean?

AI is everywhere right now. From splashy product launches to bold claims of “autonomous factories,” the conversation is moving fast—and often leaves lean practitioners wondering: Where does this fit with what we know works?

Across the lean community, respected voices are experimenting carefully while warning against the hype:

• Jamie Flinchbaugh urges transparency. If AI influences a decision or report, make it visible. Hidden AI erodes trust, while visible AI invites dialogue and learning.
• Mark Graban has shown how AI can serve as a just-in-time knowledge assistant, using his “Lean Hospitals” content to help practitioners instantly access standard work—reducing waste in the flow of knowledge without replacing critical thinking.
• The Lean Enterprise Institute (LEI) has piloted experiments with AI to demonstrate how technology can raise the floor of coaching quality, particularly for newer leaders learning scientific problem-solving.

Beyond the lean world, companies like Toyota and GE Appliances are pairing AI with lean methods:

• Toyota’s internal AI platform has saved thousands of hours of manual work, accelerating plan-do-check-act (PDCA) cycles and freeing capacity for kaizen.
• GE Appliances combines robotics and AI-driven metrology to improve flow, accuracy and safety—proof that lean cultures lead with people and purpose, while using technology to remove friction, not replace humans.

The winners won’t ask “AI or lean?” They’ll ask how to apply AI as a thought partner to enable and apply lean better—elevating strategy, execution and alignment.

Starting Point: The AI-Driven Leader Framework 

In his book The AI-Driven Leader, Geoff Woods frames AI as more than just a digital tool. He positions it as a thinking partner that leaders can interact with to improve decision-making and strategy. 

Two of Woods’ concepts are especially relevant for Lean transformations: 

1. Personas: Diverse perspectives without groupthink 

AI can adopt specific personas to evaluate plans, decisions, or strategies through different lenses. 

For example, when considering a proposed value stream design, you could prompt AI to view it as: 

• A finance leader, concerned with ROI, cash, and cost drivers. 
• A customer, focused on service levels, quality, and lead time. 
• A supply chain leader, balancing capacity, constraints, and risk. 
• Or even a board member, looking at growth and long-term positioning. 

This isn’t about “digital gemba walks.” It’s about gaining deeper insight and perspective—quickly surfacing risks and opportunities before engaging your human teams in real-world catchball. 

The result: sharper questions, better preparation, and faster alignment. 

2. Digital catchball: AI + analytics for strategic alignment 

In lean strategy deployment (hoshin kanri), catchball is the back-and-forth dialogue that refines strategies and builds alignment. 

AI enhances this process by combining data analytics with dialogue: 

• Analyzing market data, competitive intelligence and industry trends. 
• Mapping headwinds and tailwinds across your value streams. 
• Building structured tools like SWOT analyses or Porter’s Five Forces. 
• Identifying risks, blind spots, and emerging opportunities. 

Once the data is synthesized, AI can engage in catchball dialogue, challenging assumptions and refining priorities. Imagine asking: 

“Given these market headwinds, which strategic priorities align best with our long-term vision?” 

AI can even interview you—within a defined context—to uncover your goals, constraints, and ideas, then draft a summary action plan, marketing strategy, or execution roadmap for you to review and refine. 

This turns AI into a co-strategist, accelerating planning cycles while keeping humans firmly in control of decisions.

Connecting Back to Lean: Amplifying, Not Replacing

AI’s power lies not in doing the thinking for us, but in clearing the noise so leaders can focus on true problem-solving.

In our January 2025 Next Level Partners blog, we wrote:

“AI should be deployed to remove friction from decision-making, not to replace it.”

Thinking and creativity are inherently value-added. What AI can do is sift through complexity and surface patterns and free leaders to spend more time on kaizen, coaching and strategy execution—the essence of lean leadership.

Just as jidoka elevates humans by stopping the line to highlight abnormalities, AI shines a light on ambiguity and complexity, giving people the space to engage deeply and creatively.

Seven Practical Ways to Use AI in Lean Transformations

1. Strengthening problem-solving

AI can dramatically improve the discipline and speed of problem-solving by serving as a coach and reviewer:

• Reviewing A3s or 8D reports for gaps in logic, evidence and root-cause connections.
• Asking structured “5 Why” questions to test the flow of reasoning and uncover deeper causes.
• Brainstorming potential causes and breakdowns using Ishikawa (fishbone) categories, such as methods, materials, machines, manpower, measurement and environment.
• Suggesting countermeasures to test—not dictate—so teams can validate through experiments.

AI becomes an always-available sensei, pushing teams to think more deeply and improve the quality of problem-solving outputs.

2. Standard work creation and retrieval

Capturing tribal knowledge is one of the hardest parts of sustaining a lean transformation. AI and related technologies can help:

• Google Lens–style applications can digitally capture flipcharts, whiteboards or Post-It Notes directly from the gemba, instantly converting them into searchable, shareable documentation.
• This eliminates the non-value-added task of manual transcription, freeing teams to focus on experimentation and follow-up instead of paperwork.
• AI can then refine and format the captured content into clear, standardized work instructions, instantly retrievable via QR codes or mobile devices at the point of use.

This turns the messy, analog output of kaizen events into living, dynamic standards without slowing down the pace of improvement.

Anything developed by AI needs to be reviewed for accuracy and relevancy as it is dependent on the inputs but still requires someone to “trust but verify.”  We see AI as a thought-partner, not a replacement for the knowledge worker.

3. Daily management and metrics flow

AI can automatically pull data from MES, ERP and QMS systems to populate electronic boards, making it easier to track trends and spot abnormalities.

But there’s a caution here:

Decades of lean research show that direct interaction with manual boards drives ownership and engagement. When operators physically write metrics, move magnets or update visuals, they aren’t just recording numbers—they are thinking about the process, reflecting on performance, and connecting the data to what they see happening at the gemba.

This principle is supported by research on learning and cognition. Mueller and Oppenheimer’s landmark study, “The Pen Is Mightier Than the Keyboard” (2014), showed that writing by hand engages deeper thinking than typing. When operators physically interact with boards, they synthesize and interpret, not just passively observe:

“Those who write notes longhand tend to reframe and summarize concepts in their own words, engaging deeper parts of the brain,” the authors stated. “In contrast, laptop note-takers often transcribe content verbatim, resulting in shallower processing and weaker retention.”

If boards are populated automatically, they risk becoming visual noise—data “done to” operators instead of “done by” them.

Respect for people means cultivating a problem-solving culture and designing processes where the team actively engages with the data, reinforcing ownership and continuous improvement.

4. Quality at source through poka-yoke and AI

AI can be a powerful enabler of mistake-proofing, bringing traditional poka-yoke concepts into the digital era.

Computer vision systems can identify abnormalities or defects in real time, alerting operators before a defect escapes downstream or to the customer.

AI can analyze defect patterns across multiple lines or facilities, identifying systemic issues that may not be visible locally.

Predictive models can suggest design changes to fixtures or processes, integrating with engineering teams to design out sources of human error altogether.

This shifts quality from a reactive activity to a proactive system, reinforcing the lean ideal of building quality at the source rather than relying on inspection alone.

5. Predictive maintenance within total productive maintenance (TPM)

AI-powered equipment monitoring ties directly into total productive maintenance (TPM) principles. In a classic TPM framework, early steps focus on operator asset care—training team members to understand, clean and maintain their machines.

AI supports this progression by:

• Augmenting operator care routines with alerts that highlight early signs of wear or drift.
• Capturing and analyzing vibration, temperature and other signals to prevent breakdowns before they occur.
• Providing predictive insights that allow maintenance teams to move from reactive firefighting to planned interventions.

This evolution aligns with TPM’s core purpose: creating stable, reliable equipment that supports flow and continuous improvement.

AI doesn’t replace TPM fundamentals; it amplifies them by extending the reach and precision of monitoring, enabling organizations to focus on root causes rather than recurring failures.

6. Smart scheduling and PFEP alignment

The goal of production scheduling is not simply to minimize changeovers, but to serve customers effectively by producing the right product, at the right time, in the right quantity.

In a lean environment, setup time reduction (SMED) allows organizations to perform necessary changeovers quickly and predictably.

AI helps by:

• Creating data-driven production wheels, where changeovers are factored in to balance demand and flow.
• Integrating with “plan for every part” (PFEP) thinking, mapping every component’s demand, storage location, replenishment triggers and material flow.
• Recommending inventory policies that minimize shortages and overproduction, aligning production sequences with customer demand rather than internal convenience.

The result is a system that uses AI to support takt-driven scheduling, while humans focus on continuous improvement of flow and responsiveness.

7. Voice of customer and market insights for innovation 

AI’s ability to process large volumes of unstructured text makes it ideal for analyzing customer feedback, warranty claims, and complaints.

Beyond guiding problem selection, this data can feed into market research and product development, closing the loop between current performance and future innovation:

• Identifying emerging customer needs or pain points that traditional surveys may miss.
• Clustering feedback into themes that inform new product development (NPD) priorities.
• Connecting voice of customer data directly to hoshin planning, ensuring strategic objectives align with real-world customer expectations.

By integrating operational excellence with market insights, organizations can build products and services that meet demand more effectively while reducing the waste of mismatched offerings.