PlantPilot is the plant platform for everyone who builds, services and runs an industrial plant. It has three layers: a collaborative as-built twin in which OEM, EPC, engineering partners and operator work together — from the P&ID to the purchase order; a data hub that turns the sensor data you already collect into predictive maintenance and predictive operations; and a TCO optimization that computes how to run the plant at the lowest total cost per unit — energy, material, maintenance and CO₂ included.
No, that is the point. A plant is delivered by an OEM or technology provider, built by an EPC, changed by engineering offices and run by an operator — and each of them keeps its own drawings, status and data. In PlantPilot the party that is the brand on the plant (typically the OEM or main contractor) opens the twin and invites its EPC and engineering partners; the operator invites its service contractors. Everyone sees the same plant with the access they need.
Most digital twins are a digital folder: a model of the plant that one company owns. PlantPilot's twin is collaborative and transactional — partners work in it, re-engineering requests and spare-part orders run through it, and it is connected to the data hub, so the plant's structure and its live behavior sit in one place. Digital twins from equipment vendors also rarely help you buy or re-engineer parts from someone else; PlantPilot is built for exactly that.
PlantPilot was born from real deployments: a plant-service platform built with Linde (PLANTSERV®) and proven with operators like Wacker, BASF and Westfalen. mantro, the company builder behind it, turned that experience into a product every OEM, EPC and operator can use.
Yes — that is the core use case. Scanned P&IDs and documents are classified by AI, components get tagged, and the plant receives a navigable digital file. 3D models are supported but optional. Most of our reference plants are brownfield sites with decades of paper documentation.
Everything that runs on P&IDs, rotating equipment and sensors: oil & gas, petrochemical and chemical sites, hydrogen, air separation, paper, general industrial manufacturing, life science, infrastructure, renewables, mining — and increasingly data centers and other large technical facilities. The platform adapts terminology, document types and compliance requirements per industry.
No — it is the layer your CMMS is missing. A CMMS manages work orders; PlantPilot adds the collaborative twin, the partner network, the data hub and the TCO optimization on top, and connects to your existing CMMS/ERP via APIs. Purchase orders can flow straight back into your ERP.
No. PlantPilot is advisory: it reads operating data (read-only), recommends operating strategies toward proven optimal states, and your operators decide. Every recommendation is audit-logged. There is no write access to your DCS or SCADA at any time.
A proof of concept on one reference plant takes about 3 months: workshops, document upload, asset tagging, partner and vendor onboarding, and the first data-hub connection. Fleet rollouts follow plant by plant — each subsequent plant is faster because document classes, partner relations and data mappings are reused.
Whatever exists: scanned P&IDs, vendor manuals, datasheets, skid packages, spare-part lists in Excel, old maintenance records. There is no minimum quality — the AI classification is built for heterogeneous, historically grown documentation. Gaps become visible systematically instead of during the next outage.
Yes. REST APIs and standard connectors cover the common landscape: SAP (PM/MM), CMMS systems, and read-only ingestion from historians, SCADA and OEM telemetry via OPC UA into the data hub. Predictive operations work on the data you already collect — no IIoT retrofit required to start.
No. The twin, spare parts and turnaround features work entirely without sensor data. The data hub uses historian, condition-monitoring and OEM data most plants already record — usually far more than anyone analyzes. Additional metering can be added later where the TCO business case demands it.
Click a component in the drawing, see its parts and stock, request quotes from matched vendors — frame contracts are applied automatically — compare, and order. Obsolete or engineered parts go to the original OEM or EPC as a digital re-engineering request. Typical savings: around 30% on spare-part spend.
PlantPilot sends a structured re-engineering request — including drawing excerpts and specifications — to the original OEM, EPC or qualified manufacturers, who work directly in the twin. Obsolescence management also works proactively: critical single-source parts are flagged before they become a problem.
Through partner onboarding: qualification, product categories, delivery regions and engineering competences are stored once, then partners receive matching RFQs and re-engineering requests automatically. Quoting is free; a transaction fee applies only on successful orders. OEMs and EPCs can additionally open their whole installed base on the platform. See our partner page for details.
With a live readiness score per work package: parts ordered and delivered, documentation complete, permits ready, crews mobilized — with the OEM's and EPC's engineering in the same workspace. Weak spots are visible months ahead — instead of being discovered in the shutdown window at seven-figure daily cost.
The layer that collects all sensor and actuator data of the plant — from the operator's historian and SCADA, from the OEM's telemetry, from condition monitoring — read-only and mapped to the components of the twin. On top of it, pattern recognition learns how the plant behaves: which states it runs best in, when it drifts, and which asset will fail next. That is predictive maintenance and predictive operations from data you already record.
You tell PlantPilot the volume you have to produce in a given time. From the plant's own history it derives the operating strategy with the lowest total cost of ownership: energy per unit, material yield, wear and maintenance effort, downtime risk and CO₂ cost. The result is a recommended operating point and a quantified saving — typically 5–20% on energy and material per unit, plus less maintenance. In chemical plants, where energy and materials dominate input and output, these levers are significant.
Inside the TCO view. CO₂ is a cost like energy and material, and inefficiency shows up as both at once. Scope 1/2 reports per energy carrier and product carbon footprints per unit are generated from the same metered data, every figure traceable to a meter and timestamp — CSRD-ready, without a second data collection.
Pattern recognition over your own historian data identifies the operating states in which the plant ran best under comparable conditions and recommends how to reproduce them. Because recommendations are grounded in states your plant has actually achieved, they are credible and adoptable. 5–20% is the range across our reference plants; the exact figure depends on process type and energy intensity.
EU data residency, strict tenant isolation, read-only OT ingestion and a full audit trail — with an ISO 27001 / IEC 62443 certification roadmap. OT networks are never exposed: data flows one-way from historian to platform.
The operator owns the plant documentation and operating data; the OEM owns its product data. Every partner sees exactly the scope it is granted: vendors only the RFQ-relevant excerpt, engineering partners the components they work on — never your full documentation, stock levels or consumption data.
Per-plant annual licenses (Base / Pro / Enterprise) plus a 2% transaction fee on marketplace orders. OEMs and EPCs that open their installed base get platform terms per fleet. Onboarding starts with a paid proof of concept — typically amortized within the first turnaround.
The three layers compound: −30% on spare-part spend and up to −70% organization time from the shared twin, −20% unplanned downtime from predictive operations, and 5–20% lower cost per unit at the TCO optimum. Most operators reach payback within the first year — at the latest with the first turnaround prepared on the platform.
Deutsch (DE)