AUTONEX brings discovery, solution design and plant operations into one workspace. It turns operating data into OEE, equipment health and predictive-maintenance decisions, lets AI agents prepare the work, and keeps people in charge of every action. The APEX design workspace is built in.
APEX captures the customer, assesses readiness and designs the architecture. AUTONEX connects the plant data, predicts failures and losses, and turns them into approved work. Both share one sign-in, one customer context, one knowledge base and one data store.
The roadmap shows which phases are already in place, based on what exists in the platform: a saved context, connected data, analytics, decisions and coverage.
Every module reads the same plant data and customer context, so a change in one place shows up everywhere: replace a dataset and OEE, health, recommendations and the roadmap recalculate.
All APEX modules run on the same sign-in and data. AUTONEX Discover & Assess and Solution Architect use the APEX context and architecture engines directly; the full APEX workspace is one click away from every page.
Click any box to open that module. Lines show what each module hands to the next.
Each run shows its evidence: the signals, the knowledge passages and the plan. Nothing is executed until a maintenance planner approves, and every decision is recorded.
Ten factory floors on two sites with CNC machines, PLC- and DCS-controlled lines, dosing systems and belt conveyors. Maintenance is mostly reactive and OEE is not measured consistently. The programme starts with a 14-week pilot on floor 3 with 25 critical assets, then scales floor by floor.
Figures calculated by AUTONEX from the bundled Fiber Manufacturing operating dataset (14 days of production, 30 days of condition data, 12 months of work orders), as of 30 Sep 2026. They change when the customer's data is loaded.
Each finding carries its evidence (signal vs baseline and alarm limit, trend, work-order history) and a recommended action with due date and avoided downtime.
A static site behind a content delivery network, one API and one serverless function with modular capabilities, a private data store and a semantic knowledge index. No servers to manage; the plant is read, never written to.
Viewer, consultant, architect, PM and admin. Passwords are stored only as salted hashes; sessions are signed and expire. Enterprise SSO will be enabled during customer onboarding.
Agents recommend; planners approve, defer or reject with a reason. Approval creates the work order. Rejections are kept to improve future alerts.
Every decision, work-order change and data replacement is logged with user and time.
Health, RUL and diagnosis show their evidence: signal vs baseline, alarm limits, trend quality and history. Copilot answers list their sources.
Operating data, contexts and designs stay in a private, encrypted store. No secrets in the browser; the site talks only to its own API.
Plant systems are read, not written. Data is validated on upload; invalid rows are rejected and reported.
Sign in with your account. Start in the Control Tower, or open Discover & Assess to capture a new customer.