Autonomous operationsfor mission-critical industry.

An Industrial Physical AI platform: NVIDIA Cosmos 3 world models, the Wittra 4D sense mesh and AVEVA PI industrial data, bound into one operating picture — so humanoids, fleets and people work as one on the factory floor, in the warehouse, underground and beyond GPS.

With a 45-minute briefing on your use case and an NDA. We then simulate your site — same floor, same fleet, same physics — and only after the numbers hold in simulation does anything move on your floor. REAL → SIM → REAL.

Inside your perimeter. The whole stack — world models, spatial graph, fleet cognition — runs on your NVIDIA infrastructure, on-prem or air-gapped. Nothing leaves the site.

Yes — that is the point. The platform is built on measured spatial truth rather than inference: indoors, underground, between steel, in dust and darkness. When cameras go blind, the world model still holds one accurate picture of every machine and every person.

No. Deterministic planning proposes, agentic supervision approves or vetoes, and people stay on the loop. Every decision carries its evidence — no unverified model output ever moves a machine.

Physical AI is artificial intelligence that perceives, reasons and acts in the physical world — humanoids, autonomous mobile robots, forklifts, drones and whole plants — rather than only producing text or images on a screen. It is trained on sensor data, physics and real operations, and it has to be safe, because its outputs move machines around people. Systown builds Physical AI for mission-critical industry.

A world model is an AI model that learns how the physical world evolves: given what a machine sees and what it does, it predicts what happens next. That gives robots imagination — the ability to simulate, plan and check a move before making it. NVIDIA Cosmos is a family of open world foundation models for Physical AI; Systown grounds Cosmos 3 world models in measured position and industrial data so they describe your site, not a generic one.

Industrial Physical AI is Physical AI applied to mission-critical industry: manufacturing lines, warehouses, process plants, mines, ports, utilities and defence sites. It differs from consumer robotics in three ways — it must run on the operator’s own infrastructure, it must be grounded in the industrial data the plant already trusts (AVEVA PI), and it must work where cameras and GPS fail. Systown’s four products — missionGRAPH, swarmCORTEX, autonomousPI MCP and agenticOPS — are built for exactly that.

AVEVA PI (the PI System, now AVEVA PI Data Infrastructure) is the industrial real-time data platform used across process and manufacturing industries for decades. The PI Data Archive stores time-series points; the Asset Framework gives them context — sites, units, machines, attributes — as one universal namespace; the PI Web API exposes both. Systown builds on AVEVA PI: it is the memory of the plant and the training data for the plant’s own world model, and autonomousPI MCP lets an agent create, manage, analyse and forecast that data in plain language.

Spatial intelligence is knowing where everything is — every machine, vehicle, person and asset — in three dimensions and over time, and how those things relate. Cameras infer position and fail in dust, darkness and behind walls; GPS fails indoors and underground. The Wittra 4D sense mesh measures position instead, ±50 cm per the Wittra platform specification, indoors and underground. Systown lands that measured position on missionGRAPH so every robot and every world model share one spatial truth.

NVIDIA Inception is NVIDIA’s program for AI startups, giving members access to NVIDIA technology, tooling and expertise. Systown is an NVIDIA Inception Program company — an independent AI company headquartered in Stockholm that builds on NVIDIA Cosmos 3 world models and NVIDIA infrastructure. Membership is not investment or endorsement; Systown is not “backed by NVIDIA”.

NVIDIA Cosmos is an open platform of world foundation models for Physical AI: models that generate and predict physically plausible video and world states, with open weights, tokenizers and post-training tooling, running on NVIDIA silicon from Jetson at the edge to DGX in the data center. We consider it the leading world model platform for industry because it is open, omnimodal, trained at scale, and made to be post-trained on your own data — which is what Systown does: measured spatial truth and AVEVA PI context in, a world model of your site out. Systown makes Cosmos 3 potential up to 20% better on fleet efficiency and safety — measured in simulation, labelled on screen.

World models can see, hear, read and act, but none of those senses tells a machine where it is or where everyone else is once cameras go blind and GPS dies. That is the missing modality: measured position and live industrial context as a first-class input. Systown supplies it. missionGRAPH lands Wittra 4D positioning and AVEVA PI data on one live graph; the Cosmos 3 adapter delivers that measured truth into the world model in real time; swarmCORTEX turns it into one plan for the whole fleet. Vision sees. The graph knows.

autonomousPI MCP is Systown’s Model Context Protocol server over the AVEVA PI Web API — AVEVA PI Autonomous Infrastructure. Connected to Claude Fable 5.1, Claude Code, OpenAI ChatGPT, Google Gemini, Astra 6 or any MCP client, an agent can browse any Asset Framework, find and read any of an unlimited number of PI points, provision whole AF-structured estates in one call, and forecast any point zero-shot with Google TimesFM. Read-only by default, on your infrastructure.

agenticOPS is Agentic Industrial Operations: domain-fluent AI agents that reason over live AVEVA PI data and the missionGRAPH world picture, run multi-step workflows — alarm to work order, forecast to setpoint proposal, blocked aisle to re-route — optimize energy, throughput and maintenance windows against your own history, and act only within hard guardrails. Deterministic planning proposes, agents approve or veto with a stated reason, people stay on the loop, and every decision is on the record.

Systown is an NVIDIA Inception Program company, and the platform builds on NVIDIA Cosmos 3 world models. Systown is an independent AI company headquartered in Stockholm.