Lights-out Factory

A Lights-Out Factory is a plant where production runs without human intervention: machines run, monitor and correct themselves while the lights are literally off. Almost no plant works this way end to end today, so in practice it is a direction rather than a destination. Factories get there by moving tasks up a ladder of autonomy, from AI that advises people, to AI that works alongside them, to loops that close on their own.

LTTS draws the line at one question: who closes the loop? A smart factory is connected and data-rich, but a person still reads the dashboard and makes the call. In a Lights-Out Factory, the loop closes without waiting for anyone. Smart manufacturing gives you visibility. Autonomous manufacturing gives you a decision. LTTS moves plants from one to the other, task by task.

LTTS uses AI for predictive maintenance rather than reactive repair. The MECHA accelerator tracks temperature, speed, feed, energy and vibration to predict machine health, so maintenance is scheduled instead of emergency. AgeOps agents go further, monitoring robot health and acting on faults to speed up root-cause analysis. A large plant loses around 27 hours a month to unplanned downtime.

LTTS applies the Lights-Out Factory framework wherever a process can be modelled and measured. The deepest proof is automotive: body-in-white, paint shop, spot welding and final assembly. LTTS also works in industrial machinery through equipment digital twins for rotating assets, in composites through AI-assisted impact testing, and across logistics and utilities. The framework is sector-neutral; the accelerators are tuned to the sector.

LTTS targets the numbers a plant leader owns: less unplanned downtime, shorter cycle times, fewer defects found late, and capital committed only after it is proven in simulation. In one engagement with a North America-based automotive OEM, LTTS delivered around 30% improvement in collaboration, visualisation speed and data consistency at proof-of-concept stage.

LTTS builds Lights-Out Factory operations one loop at a time. The engagement runs in four stages: Assess, Build, Operate and Sustain. LTTS scores every task on your line against three modes of autonomy, builds a validated digital twin of the real production system, then deploys AI accelerators at the level each task has earned. A task only moves up when the evidence supports it. Autonomy is earned, not installed.

LTTS treats the digital twin as the foundation everything else runs on. It is a validated virtual model of your plant where AI learns your actual process, a change is tested before it costs anything, and a decision is proven before it reaches a real machine. Without a digital twin, autonomous manufacturing has nowhere safe to fail.

LTTS engineers across the mixed-vendor environments real factories run on, not a single preferred stack. For a North America-based automotive OEM, LTTS converted legacy simulation and virtual-commissioning models into one live environment, integrating PLC, HMI and robot programs for real-time virtual commissioning. Integration is part of the offering, not an exception to it.

LTTS builds its Lights-Out Factory framework on NVIDIA Omniverse and NVIDIA Isaac. Omniverse is the shared environment where simulation models, plant layouts and live control systems come together as one digital twin. Isaac Sim and Isaac Lab are where LTTS trains and validates robotics, from navigation to bin-picking, before anything reaches real hardware. LTTS presented this work jointly with NVIDIA at GTC 2026.