Your next major overhaul has 7,000 work packages, 28 contractors arriving in different weeks, 1,400 spare parts with lead times ranging from 2 days to 14 months, and exactly 21 days of outage window. Miss one critical-path item and you slip the return-to-service date. Slip the return-to-service date and your CFO is looking at $400,000 of lost generation per day plus emergency-rate premiums on every contractor still on-site. Most plants still build this schedule in spreadsheets and Microsoft Project files that nobody updates after day three. The Shutdown Management Agent changes that completely. Type your overhaul scope, get a fully sequenced critical path in seconds — every parallel workstream identified, every long-lead spare flagged, every contractor's arrival window mapped against access constraints. One large utility cut their last major outage by 4 days using exactly this approach. Register for the event to plan your next overhaul live with our team.
MAY 12, 2026 5:30 PM EST , Orlando
Upcoming OxMaint AI Live Webinar — Shutdown Management Agent Live Demo
Live session for outage managers, maintenance heads, plant directors, and reliability leaders evaluating on-prem shutdown AI. The Shutdown Management Agent will be running on real plant overhaul data — sequencing 7,000+ work packages, identifying parallel workstreams, flagging long-lead spares, and mapping contractor arrival windows. Bring your next overhaul scope to the booth. Watch the agent build your critical path in front of you. Hands-on time on the actual hardware. Walk out with a quote and an order date. Pilot to fully running in 6 to 12 weeks.
The On-Prem Server Stack — Your Outage Planning Engine, In a Box
Outage planning data is sensitive. Contractor rates, scope inspection findings, OEM-confidential procedures — the kind of data your IT and procurement teams want kept inside your perimeter. Three pieces of hardware sit in your control room, never connect to the internet, and crunch your full overhaul scope in seconds instead of weeks. Register for the event to see the full setup running live.
AT THE PLANT FLOOR
NVIDIA Jetson AGX Orin The Listener
JobPulls condition data + inspection findings from your CMMS
SpeedReal-time scope updates as inspections come in
WherePlugs into your existing CMMS and inventory
FormSmall box, mounts on a rail
What it does: Feeds the agent your live scope register — every inspection finding, every condition alert, every deferred PM, every spare on order. Keeps the schedule current without manual updates.
10K+WORK PACKAGES TRACKED
CONTROL ROOM · THE BRAIN
RTX PRO 6000 Blackwell The Planner
JobSequences the full critical path for your overhaul
SpeedFull overhaul plan in under 30 seconds
WhereSits in your control room behind your firewall
FormTower computer, fits under a desk
What it does: The actual planning engine. Optimizes 5,000-10,000 work packages against your contractor windows, spare lead times, access constraints, and permit-to-work rules. Builds the Gantt your team executes.
<30 secFULL SCHEDULE BUILD
ENTERPRISE · MULTI-PLANT FLEET
NVIDIA DGX Station GB300 Ultra · Fleet Brain
JobCoordinates overhauls across multiple plants
SpeedCross-plant resource leveling in hours
WhereSits at corporate HQ in a server rack
FormRack-mounted, 24/7 enterprise grade
What it does: Stops two of your plants from booking the same OEM team in the same week. Coordinates spare parts inventory across the fleet. Builds the rolling 5-year overhaul calendar for the executive team.
25+PLANTS · ONE BRAIN
100%
Stays inside your building · never goes online
$0/mo
No subscription · buy once, own forever
Offline
Works fully air-gapped if your team requires it
Yours
Source code included · modify it freely
The Critical Path — What an AI-Built Overhaul Schedule Actually Looks Like
Below is a real 21-day major overhaul schedule built by the Shutdown Management Agent. Five workstreams running in parallel, every dependency mapped, every long-lead spare locked in months ahead. The gold bars are the critical path — anything that slips here slips your return-to-service. The other bars run in parallel, hidden inside the critical path's shadow. Register for the event to walk through this Gantt live with our team.
21-Day Major Overhaul · Critical Path Visualization
Three Real Overhaul Problems — How the Hardware Solves Each One
Three real outage scenarios from real plants — each one walked through in plain language. Same pattern every time: plant floor listener → control room planner → contractor portal + work order. Register for the event to see these scenarios replayed live.
SCENARIO 01
"We're 90 days from the major overhaul. Is our scope locked or are we walking into a 4-day surprise?"
THE PROBLEM
90 days from a major overhaul. Your team has a scope spreadsheet with 3,200 line items. Pre-outage inspections last week added 480 more findings — boiler tube thinning, valve seat erosion, bearing condition issues. Some of those findings need long-lead spares (HP turbine blades — 14-month lead time). Some need OEM team time. If even three of those get missed in scope, you're looking at 4 days of unplanned discovery work mid-outage. The maintenance director needs to know which findings are critical-path before tomorrow's review meeting.
HOW THE HARDWARE SOLVES IT
The Listener (Jetson)
Pulls all 480 inspection findings into the agent automatically — no manual data entry.
The Planner (RTX)
Categorizes each finding: critical path (47), parallel workstream (185), defer to next outage (248). Flags 14 long-lead spares that must be ordered this week. Builds revised Gantt in 28 seconds.
Outage Plan
Auto-creates: 47 work packages flagged "scope-locked." 14 purchase requests sent. Updated Gantt shows new RTS still hits original date. Director walks into the meeting with answers.
THE RESULT
Outage came in 2 days under schedule. Zero discovery surprises. $800K of avoided overrun costs. RTS hit on time.
SCENARIO 02
"Two contractors just told us they need the same equipment hatch on the same day. Now what?"
THE PROBLEM
Day 9 of the outage. The OEM turbine team needs the equipment hatch open at 6 AM tomorrow to bring in the rotor stand. The boiler contractor needs the same hatch open at 6 AM tomorrow for tube bundle delivery. Both are blocking critical-path work. Resolve it badly and one of them sits idle at $4,800/hour while the other works. Resolve it without thinking and you create a downstream conflict on day 11. The outage manager needs an answer in the next 2 hours.
HOW THE HARDWARE SOLVES IT
The Listener (Jetson)
Sees both contractor work packages flagged as critical path with overlapping access requirements.
The Planner (RTX)
Tests 12 sequencing options. Best plan: turbine team gets hatch 6-9 AM (rotor delivery), boiler team gets it 9 AM-noon (tube bundle), then both teams resume parallel work afternoon. RTS unaffected. Confidence: high.
Contractor Portal
Both contractor leads get push notification on their phones with the new sequence and access window. Auto-updates the daily plan. No 2-hour control room meeting required.
THE RESULT
Conflict resolved in 4 minutes. Zero contractor idle time. $60K saved that day. Critical path intact.
SCENARIO 03
"Day 14: We just found a generator wedge tightness problem. Can we still hit RTS on day 21?"
THE PROBLEM
Day 14 of a 21-day overhaul. Generator wedge tightness inspection found 3 wedges loose. Repair adds an estimated 3 days if done sequentially. Skip it and you risk a forced outage in 3 months. Plant director needs to call the dispatch desk in the next 30 minutes — extend RTS by 3 days, or add weekend OT and pull other tasks earlier? Either decision costs hundreds of thousands; the wrong one costs millions.
HOW THE HARDWARE SOLVES IT
The Listener (Jetson)
Captures the inspection finding plus current resource availability — what crews are free, what's still on the critical path.
The Planner (RTX)
Runs 200 schedule scenarios. Best plan: split wedge repair into two 12-hour shifts running parallel with turbine box-up, pull HX testing earlier. RTS slides only 16 hours, not 3 days. Saves $1.2M vs full slip.
Outage Plan
Auto-updates the Gantt. Pushes new shifts to crew lead phones. Generates dispatch notice: "RTS slip 16 hours, day 21 → day 22 morning."
THE RESULT
Wedges repaired. Unit returned 16 hours late instead of 3 days late. $1.2M saved vs original sequential plan. Forced-outage risk eliminated.
$2M+
Combined savings across the three scenarios — scope discipline, contractor coordination, and live re-planning. The hardware pays for itself in the first overhaul where it prevents a multi-day overrun.
Why This Matters — One Day Saved Pays for the Stack
Without an AI planner, every overhaul is a paper plan that breaks on contact with reality. With one, your critical path stays visible from day one to day twenty-one, every parallel opportunity gets used, and discovery work doesn't blow up your RTS. The numbers below are what real overhauls look like.
25-35%
Reduction in total outage calendar days when parallel workstreams are properly sequenced
$400K
Cost of a single day of unplanned overhaul extension — lost generation plus contractor premium rates
5-10K
Work packages in a single major overhaul — far too many for spreadsheets to track
15-40
Specialist contractor teams running in parallel during a major overhaul
6-12 wk
From hardware delivery to live planner running on your next overhaul
$0/mo
No subscription, ever. Buy the system once. Own it forever. Source code included.
May 12 · 5:30 PM EST · Orlando · Hands-On
Bring Your Next Overhaul Scope. Watch the Agent Build Your Critical Path.
Walk in with your draft scope register — or even just your top 200 work packages. Watch the agent sequence the critical path, identify parallel workstreams, and flag long-lead spares in front of you. Get a real Gantt, real durations, real contractor windows — all in 30 seconds. Walk out with a quote and an order date. Pilot to fully running in 6 to 12 weeks.
What You Get — Everything In One Box, Yours Forever
A pre-configured server arrives at your control room. The Shutdown Management Agent is already loaded — trained on hundreds of historical overhauls from comparable thermal plants. Plug it in, point it at your existing CMMS, watch it baseline against your scope register within weeks.
Buy Once, Own Forever
No monthly fees. No per-overhaul charges. No annual price increases. Pay one price up front.
Your Data Stays Home
Contractor rates, OEM procedures, scope findings — all stays inside your perimeter. IT and procurement teams approve fast.
Source Code Included
Add your own scope templates, custom workflows, or build new agents. Not locked in to us forever.
Ready Out of the Box
Critical-path engine + contractor portal + spare lead-time tracker — all pre-loaded. No 6-month integration.
Frequently Asked Questions
How fast can the Shutdown Management Agent run our next overhaul?
From the day you sign the order to the day your team is using it daily is usually 14 to 22 weeks. The hardware arrives in 4 to 6 weeks already pre-set-up. Once on-site, our team connects it to your existing CMMS, inventory system, and contractor records. The agent then trains on your historical overhaul data. Your team is using it for outage planning in 6 to 12 weeks. For your first AI-planned overhaul, we recommend starting scope ingestion 90 to 180 days before the outage window.
Will it work with the CMMS we already have?
Yes. The agent reads from any major CMMS — SAP PM, Maximo, eMaint, Fiix, Infor EAM, OxMaint's own platform, or custom in-house systems. Same goes for inventory: integrates with SAP MM, Oracle, JD Edwards, and others. Contractor data can come from spreadsheets, ERP, or be entered directly. The Jetson edge gateway handles the protocol conversion regardless of brand.
Can it really sequence 5,000+ work packages, or just produce a high-level summary?
Real sequencing — every individual work package, with predecessor logic, resource assignment, and access constraints. The agent is designed for high-volume overhaul events: 5,000 to 10,000 discrete work packages within a single outage without performance degradation. Filter by discipline, system, contractor, or critical path so the outage manager can navigate without reviewing every task individually.
What about contractors who don't want to use yet another portal?
The contractor portal is intentionally minimal. Each contractor lead sees only their own assigned work packages, predecessor task status, and access requirements — not the full plant CMMS. Mobile-first design means most contractors complete their daily updates in under 60 seconds on their phone. No training required for most teams. For contractors who refuse to use any portal, the outage manager can update on their behalf — but in practice, most pick it up by day two.
What does "you own it" really mean? Are there hidden costs?
It means exactly that. You pay one price up front for the hardware, software, and source code. No monthly fees. No per-user charges. No annual increases. The only optional future costs are entirely your choice: support contracts if you want our help, custom feature work if you want something specific built, or hardware refresh whenever you decide. You can run the system forever without paying us anything more.
May 12 · 5:30 PM EST · Hands-On Hardware
Lock Your Spot. Stop Your Next Overhaul From Slipping.
Walk in with your overhaul scope. Watch the agent sequence it. Talk to the engineers. Leave with a quote and an order date. Pilot to fully running in 6 to 12 weeks. Buy it once, own it forever — no monthly fees, ever.
