Digital Kaizen: How VR, AR and AI Are Revolutionising Gemba Walks

Introduction

Gemba walks are a pillar of lean culture. They encourage leaders to go to the place where value is created—whether a factory floor, a hospital ward or an office—observe processes, listen to frontline staff, ask insightful questions and identify waste. Traditionally, gemba walks were purely physical: managers walked the line, carried clipboards and relied on their senses and conversations to uncover opportunities. In today’s complex world, organisations operate across multiple sites and often have remote knowledge workers. Travel costs, schedule conflicts and global supply networks make it difficult for leaders to physically observe every process. Meanwhile, digital technologies are transforming how we work. Virtual reality (VR), augmented reality (AR), artificial intelligence (AI) and the Internet of Things (IoT) can extend our ability to see and understand processes. This article explores how these technologies are changing gemba walks and how they can be used to enhance—rather than replace—the human side of continuous improvement. Drawing from lean principles and research on technology’s role in lean , we show that digital tools are effective only when they support respect for people and problem-solving.

Why Traditional Gemba Walks Still Matter

The purpose of a gemba walk is not to conduct an audit but to support, coach and learn. When leaders visit the place where work happens, they see reality as it is. They observe flows of materials and information, notice deviations from standard work, ask frontline employees what impedes them and encourage improvement. In-person walks build trust and empathy—leaders learn to ask open-ended questions, listen without judging and empower workers. The discipline of regular gemba visits keeps improvement at the forefront and prevents complacency. Traditional gemba walks also emphasise seeing with the eyes, ears, nose and hands. You notice vibrations, smells and sounds that indicate problems. You feel the fatigue of lifting heavy parts and sense how long people wait. These experiences can be lost when you rely only on data.

However, manual gemba walks have limitations. When plants are spread across continents, it is impractical for executives to visit every site frequently. Some facilities have dangerous areas that require special training, limiting who can observe them. Knowledge work takes place in digital systems like enterprise resource planning (ERP) software, not on the shop floor. Most importantly, by the time you observe an issue, it may already have led to scrap or delays. Digital technologies can complement physical gemba walks by providing real-time data, remote access and advanced visualisations.

VR: Immersive Process Observation

Virtual reality can transport leaders to a production line without requiring travel. VR headsets play high-resolution 360° videos or simulations of the plant. Leaders can “walk” around a virtual representation of a line, pause the view, zoom in on details and see processes from angles that may be unsafe or difficult to access in person. For example, a chemical plant might use VR to simulate a mixing process that occurs inside a sealed vessel. VR enables managers to observe internal flow, identify dead zones and ask operators questions during the simulation. In aerospace, VR is used to rehearse maintenance procedures on aircraft where access is limited. These immersive experiences train managers to see waste, such as unnecessary movement or waiting, and help them better understand the complexity of processes.

VR is also a powerful training tool. New supervisors can practice gemba walks in a controlled environment. They can learn to identify different types of waste—motion, overproduction, waiting, defects, over-processing, inventory, transportation and skills misuse—before they enter the real factory. VR scenarios can be customised to show poor workplace organisation, safety hazards and flow disruptions. Trainees receive immediate feedback, which accelerates learning. Additionally, VR is valuable for scenario planning. Engineers and managers can import digital twins of equipment, rearrange layouts and simulate the impact on flow and ergonomics. For example, a lean team might virtually test three different cell configurations, adjusting takt times and worker placement to maximise flow. This reduces the time and cost of physical pilots.

AR: Enhancing Reality with Data

Augmented reality overlays digital information onto the real world. Unlike VR, which immerses you in a simulated environment, AR keeps you in the physical environment but adds contextual information. AR headsets, smart glasses or smartphone apps can display real-time data, instructions or alerts as you walk the floor. During an AR-assisted gemba walk, a manager could see cycle times, yield rates, inventory levels and energy consumption superimposed on each machine or workstation. Colour-coded indicators may highlight machines with frequent breakdowns or processes with long waiting times. The manager can take pictures or videos, annotate them and link them to digital value stream maps.

For operators, AR delivers standardised work instructions and reduces cognitive load. A technician performing a changeover sees a step-by-step sequence projected onto the machine. Sensors confirm each step; if a tool is forgotten or a screw is tightened incorrectly, the system alerts the operator. Similarly, maintenance personnel can access manuals, exploded diagrams and remote expert support without leaving the equipment. By reducing errors and variation, AR aligns with lean’s emphasis on building quality into the process rather than inspecting defects downstream. It also supports just-in-time by ensuring changeovers are quick and consistent.

AI: Insight Beyond Human Perception

Artificial intelligence analyses large datasets to identify patterns and predict future events. In a gemba context, AI can process data from sensors, machines and IT systems to highlight abnormal conditions, trends and correlations. Predictive maintenance algorithms alert teams when vibration patterns indicate an impending bearing failure. Process mining algorithms reconstruct the actual sequence of operations from time-stamped events, revealing bottlenecks and deviations. Machine learning can detect subtle relationships, such as how humidity levels affect scrap rates or how micro-stoppages accumulate to cause downtime. AI-driven dashboards allow leaders to ask natural-language questions—“Show me the average cycle time for line 5 last month,” or “Predict how a one-minute increase in changeover time will impact daily output.” These capabilities extend human cognition but do not replace it.

Generative AI tools like ChatGPT are particularly promising for lean documentation and knowledge sharing. They can draft standard work, A3 reports and training materials based on prompts. For instance, an engineer might ask ChatGPT to summarise the steps to clean a heat exchanger and provide safety precautions. The AI drafts a description, which a subject matter expert reviews and edits. This saves time and ensures consistency across documents. AI can also answer common lean questions from new hires, reducing training time. However, AI models must be used carefully. They may hallucinate or provide outdated information. Users should treat AI outputs as suggestions and validate them with experienced colleagues .

Building a Digital Gemba Strategy

Digital tools are only as effective as the strategy guiding them. Organisations should take a structured approach:

1. Clarify objectives. Determine why you are adopting digital tools. Do you want to increase the frequency of gemba walks, involve remote experts, train new leaders or visualise complex flows? Clear objectives prevent technology from becoming a distraction.
2. Map the current state. Begin with traditional value stream mapping: define customer value, map the value stream, create flow, establish pull and pursue perfection . Use this map to identify processes that could benefit from digital enhancement.
3. Select pilots. Choose processes or lines that are stable and have engaged teams. Introduce one technology at a time. For example, pilot AR glasses on changeovers or remote VR walkthroughs for a specific line. Collect baseline data and measure improvement in safety, quality, delivery, cost and morale.
4. Train users. Provide hands-on training for VR, AR and AI tools. Teach leaders to ask the right questions and listen attentively during digital gemba sessions. Train operators to use AR interfaces and trust AI alerts. Emphasise that technology complements, not replaces, human judgement.
5. Integrate with continuous improvement. Digital observations should feed into kaizen logs, suggestion systems and A3 problem-solving. Link AI alerts to existing escalation processes. Celebrate improvements and share lessons learned.
6. Review and adapt. After each digital gemba, conduct reflections. What did you learn? What worked? What didn’t? How can the technology be improved? Adjust hardware, software and training accordingly.

Overcoming Challenges

Adopting digital gemba practices presents challenges. One risk is data overload. Dashboards may display hundreds of metrics, but only a few matter for improvement. Lean practitioners must prioritise key performance indicators that reflect customer value and lean goals . Another challenge is loss of context. Remote observations may miss subtle cues—body language, team dynamics, machine sounds. Leaders should schedule periodic physical visits to maintain personal connections and holistic understanding. Data quality is critical; inaccurate sensors or poorly maintained databases can mislead. IoT and AI require investment and robust cybersecurity. Finally, people may resist change. Operators may fear surveillance or job loss. Leaders must communicate the purpose of digital tools—empowerment, safety and improvement—and involve employees in design and implementation.

Conclusion

Digital gemba is not about replacing humans with technology; it is about extending human capability. VR, AR and AI provide new ways to see processes, engage remote teams, analyse data and accelerate problem-solving. However, the success of digital gemba depends on a foundation of lean thinking: defining value, eliminating waste, creating flow and respecting people. As researchers warn, automation and AI do not automatically eliminate problems; they can amplify waste if misapplied . By integrating digital tools with human observation and problem-solving, organisations can make gemba walks more frequent, inclusive and effective, driving continuous improvement in an increasingly complex world.