Prometheus Group Bets on AI to Close the Maintenance Planning Gap in SAP Plants

For maintenance planners working in SAP Plant Maintenance, the weekly scheduling cycle has long been a grind of exports, spreadsheets, and rework. Prometheus Group argues that this cycle is now solvable. The enterprise asset management (EAM) vendor recently detailed how its GWOS-AI planning and scheduling solution compresses work that takes seven to nine hours in a native ERP system into a comparable result produced in roughly three minutes.

The company described a familiar scenario: a planner spends two days manually moving hundreds of work orders in SAP, rebuilding the schedule in Excel, and hand-delivering printouts, only to see half of it invalidated by Monday morning. That is because a single planner may be responsible for more than 1,000 active work orders at any given time.

A Workforce Problem

Prometheus Group framed the urgency in terms of workforce turnover. Before 2020, average tenure at manufacturing organizations was approximately seven years. Recent industry data cited in the post suggests average company tenure has dropped to around three years, with time in the same role closer to nine months. Learning a single ERP system such as SAP can require a full six-week training program before a planner is functional, which means some employees leave before they finish onboarding.

That erosion of institutional knowledge shows up in work orders released without materials, inaccurate job durations, and inconsistent schedules across sites. Prometheus shared an example of a chemical company whose planners had abandoned SAP scheduling entirely and were running operations out of Excel, leaving no reliable source of truth for new planners or contractors.

The answer lies in a human confirmation-required model. The AI reads notifications, suggests failure modes, pre-populates operations and durations, and can build a full weekly schedule from a plain-language prompt. Nothing writes back to the ERP until a human approves it. GWOS-AI connects directly to SAP through live, bidirectional integration, with support for both SAP ECC and SAP S/4HANA, so that organizations in transition can maintain a single scheduling interface across versions.

Part of a Broader Vertical AI Strategy

In another example on AI in oil and gas Prometheus distinguished horizontal AI, which only suggests, from vertical AI built into maintenance workflows. The company pointed to MDaaS-AI for cleansing material master data against ERP structures and RapidAPM for prioritized asset health alerts, citing offshore operators averaging around 27 days of unplanned downtime a year at a cost on the order of $38 million. Prometheus Group has also been expanding the platform through acquisitions, including the frontline-worker platform Webalo in January 2026 and the procurement specialist Tamarack in April 2026, both of which are focused on ERP integration with SAP.

The timing aligns with where SAP customers say they are heading. SAPinsider’s Technology Leaders’ Strategic Agenda for 2026 benchmark research found that 70% of respondents identified increasing operational efficiency and reducing costs as their top priority, with 40% targeting intelligent automation in core ERP processes. Yet SAPinsider’s research on AI in enterprise asset management found that only 7% of organizations have completed AI and machine learning deployments in EAM, even as many plan to do so within two years. That gap between ambition and execution is precisely the market Prometheus Group is targeting.

What This Means for SAPinsiders

AI-assisted scheduling is now a credible answer to the planner skills gap, not a future concept. With planner tenure collapsing and SAP PM onboarding measured in weeks, maintenance leaders and ERP program managers should evaluate tools like GWOS-AI that embed best practices and keep SAP as the single source of truth. The first step is auditing where scheduling actually happens today. If the answer is Excel, the organization already has a governance problem worth fixing.

Human-in-the-loop design should be the baseline requirement for AI touching SAP data. The “confirmation required” model matters because AI-generated schedules write back to SAP PM work orders, capacity plans, and assignments. CIOs and enterprise architects evaluating AI for maintenance should insist on simulation modes, approval gates, and bidirectional integration with SAP ECC and SAP S/4HANA rather than exports that create parallel data sources.

EAM AI adoption will stall without clean master data, so sequencing matters. SAPinsider research shows most organizations have not yet operationalized AI in asset management, and Prometheus Group’s own positioning makes master data the foundation. SAP teams should assess the quality of material masters and functional locations, potentially using solutions such as MDaaS-AI or SAP Master Data Governance, before layering predictive tools on top. Clean data first, then automation.