How can AI be integrated into operations without causing disruption?
Yuriy: The key to initial deployments is gaining trust and providing immediate value to the floor personnel. Nondisruptive AI assistance – seamlessly reading sensor telemetry and surfacing disparate data – should fundamentally make jobs on the floor easier. With the AI handling the immediate cognitive load, the role of experienced personnel shifts from reactively firefighting to keep the line running, to proactively optimising the process for less waste and higher throughput. Running the system in this assistive mode also gives us the time to perfect the knowledge base. In my experience, true tacit knowledge cannot be extracted through forced, scheduled interviews; it is only captured organically through shoulder-to-shoulder human-machine collaboration on the factory floor. This takes time. Once that knowledge base is solidified, the AI can safely transition into more autonomous, proactive work. Ultimately, integrating AI without disruption requires more than just deep technical expertise. It requires deep domain expertise and a commitment to gradual, evolutionary change – perfectly aligning with the Japanese tradition of Kaizen, or continuous improvement.
Premkumar: Successful integration starts with respect for operations. You must deploy AI incrementally, alongside existing workflows, beginning with decision support and progressing to automation. Use of digital twins and simulations will enable you to validate behaviour before production rollout. And youmust train and engage operators early – trust is as important as accuracy. When AI is introduced as an operator ally, not a black box, adoption accelerates naturally.
“Industrial AI operates in mission critical, physical environments – not digital sandboxes”
Premkumar Balasubramanian Chief Technology Officer and Head of AI Hitachi Digital Services
82 May 2026