Maneva launches Orchestration, mapping every AI agent into one self-improving smart factory
Orchestration is the bridge that turns any factory floor into a single intelligent platform
SACRAMENTO, Calif.– August 2026 – Maneva, the vision AI solution for manufacturers, announces Orchestration, an AI brain for factory floor management that gets smarter with every shift. Large manufacturers lose an estimated $1.4 trillion a year due to unplanned downtime, equivalent to the entire GDP of a mid-sized industrial economy. More than half of manufacturers were hit by unplanned downtime last year, according to Supply & Demand Executive, with incidents costing an average of $400,000 an hour, and in the worst cases, up to $1.7 million an hour.
Orchestration is built on four layers of intelligence, perception, understanding, execution, and knowledge. The platform turns a network of independent factories and AI agents into one coordinated manufacturing ecosystem that automatically monitors, reasons, intervenes, and reports. Maneva’s approach to connecting the factory floor is to leverage existing infrastructure with best-in-class software, giving you the ability to grow into a smart factory, solving one problem at a time, stair-stepping up to eventually a connected factory.
Orchestration builds on top of Maneva's Video-to-Action (VITA) and AI Line Supervisor (ALIS) AI models, which inspect every unit on a line for defects and downtime risks, and watch the floor for safety and productivity issues. Both run on Maneva's proprietary reinforcement learning, deployed on NVIDIA edge technology that connects to existing factory cameras, with no expensive infrastructure changes required.
Global to ground floor management
Log in as a CEO, and you can see the entire factory network. Log in as a plant manager, and it's your floor. Log in as an operator, and it's your line.
Orchestration renders every plant Maneva serves on a single map. Designated users can select a specific facility and view live floor plans showing exactly where each VITA and ALIS agent is deployed, which area it's monitoring, and whether any faults or KPIs need attention. No more switching between separate systems for quality, safety, productivity, and machine status. Zoom into any AI agent inside Orchestration and see its live feed, full metrics history, and controls, all in one place, all in real time. From the entire operation down to a single conveyor belt or cook table, manufacturers are just a few clicks away from action and improving throughput.
Key Orchestration senses
Perception
Global view: Map every facility in a manufacturer network on a single global map, with a red or green status showing which plants need attention
Live floor plan: View the live floor of any facility, showing every VITA and ALIS agent by zone and name
Live video feed: Stream a live video feed from every agent on the floor, so users can observe exactly what each agent sees, in real time
Understanding
Agent dashboards: Showcase each agents’ metrics with uptime, defect rates, safety incidents, and alert history
Root cause analysis: Gain complete visibility into the line by cross-referencing every agent in real time, so when something breaks, you can pinpoint the upstream that change that caused it
Execution
Automated intervention: Connect directly to factory hardware to allow the platform to slow a conveyor, trigger a safety stop, or send an alert without human intervention
OEE tracking: Track Overall Equipment Effectiveness at the production line level, comparing product in-feed to output, as well as downtime and productivity tracking, to show exactly where losses occur and how to mitigate them in the future
Agent configuration: Adjust agent thresholds inside the platform, so a false alarm gets fixed in seconds instead of sitting in an engineering queue
Knowledge
Ask the floor: Talk to the factory through Kaizen, Maneva’s AI chat that answers operational questions, reasoning from live plant and agent data
On-demand reports: Turn weeks of paperwork into a single request. Generate shift, safety, and output reports instantly, by line, facility, or date range
Results from early deployment
Maneva has already shipped its Orchestration Platform to early design partners, with several more deployments actively in the pipeline. In one of its initial deployments, Maneva has more than 50 VITA and ALIS agents running across a single customer's facility. Together, they cover the entire production process from end to end, spanning raw material ingestion, machine downtime monitoring, product manufacturing, health and safety, and shipping.
The result is the first signal of a self-improving factory. Rather than operating a set of discrete tools with disparate workflows, customers are beginning to see capabilities that include self-diagnosed line speed optimization, autonomous throughput optimization, adaptive shift staffing, and self-initiated work orders.
“Before Orchestration, a plant manager might find out about a problem an hour after it started, or a day, or not until a customer did. Now they know the moment it happens, and they know why, cutting downtime, catching dangerous defects before they reach a customer, and keeping workers safer on the line,” said Rae Jeong, Maneva’s Co-Founder and CEO. "Every incident observed by Orchestration makes a facility smarter. Maneva is the difference between a factory that just collects data and one that can think and improve itself."
About Maneva
Founded in 2021 and headquartered in Delaware, Maneva is the physical AI solution provider specializing in video-to-action intelligence for manufacturing. Its software-first platform integrates with existing factory camera systems to understand what's happening on the production floor in real time and communicate insights to mitigate risks and improve workflows. Maneva’s customers span the confectionery, wood, glass, steel, meat processing, food and beverage, pharmaceutical, cosmetics, and distribution industries.
Leadership
Rae is the Co-Founder and CEO of Maneva, where he leads the company’s mission to bring real AI intelligence to the factory floor. Before working in AI, Rae was a welder in Alberta, where long hours in heavy industry gave him a lasting respect for the skill and grit it takes to keep the global economy running. Those early experiences shaped his belief that the workforce deserves better tools designed to assist people as they carry the weight of industry into a new frontier. Previously a Research Engineer at Google DeepMind, Rae worked on robotics and deep reinforcement learning, specializing in dexterous manipulation.
Kelvin is the Co-Founder and CTO of Maneva, where he leads the development of the company’s patent-pending VITA and ALIS technologies. He previously worked at Magna International as a Research and Development Engineer, focusing on next-generation perception technologies involving computer vision, radio frequency, and deep learning for automotive applications. Kelvin holds a degree in Mechatronics Engineering from the University of Waterloo, a background that provides him with a deep understanding of both mechanical systems and embedded AI.
