Automating Annual Works Planning: How Yarra Trams Turned Maximo Data into Forward Forecasts

Automating Annual Works Planning: How Yarra Trams Turned Maximo Data into Forward Forecasts

When Yarra Trams approached SAS-AM to streamline their annual works planning process, they had a common problem: mountains of maintenance rule data sitting in their IBM Maximo system, but no efficient way to translate that into meaningful forward forecasts. What should have been a straightforward planning exercise was consuming weeks of manual effort each year, with multiple spreadsheets, version control issues, and limited confidence in the outputs.

The solution we developed demonstrates how the right analytical approach can transform existing data into a genuine planning asset — one that saves time, reduces errors, and provides the visibility leadership needs to make informed decisions.

The Challenge: From Rules to Reality

Yarra Trams operates one of the world’s largest tram networks, maintaining hundreds of vehicles across Melbourne’s extensive network. Their maintenance regime follows a complex web of preventive maintenance (PM) rules stored in IBM Maximo. These rules dictate when specific tasks must occur based on time intervals, operating hours, or distance travelled.

The annual works planning process required translating these rules into a three-year forward forecast — essentially predicting when every maintenance task would fall due for every asset. Previously, this involved extracting raw data from Maximo, manually cross-referencing PM rules against asset registers in spreadsheets, and building forecast models cell by cell. The process typically consumed four to six weeks of senior planner time and still left questions about accuracy and completeness.

Understanding the Data Architecture

Before automation could occur, we needed to understand what Yarra Trams actually had. Their Maximo instance contained detailed PM rule structures — the “what” and “when” of maintenance — but these rules alone don’t generate forecasts. They’re templates, not schedules.

The critical missing piece was connecting three data streams. First, the PM rule definitions themselves — task descriptions, intervals, trigger types, and applicable asset classes. Second, the asset register — which specific trams, bogies, and components each rule applies to. Third, operational data — current meter readings, last completion dates, and upcoming compliance deadlines that determine when each task next falls due.

Individually, each dataset was well maintained. The challenge was that no existing process joined them together programmatically to produce a time-phased forecast.

The Solution: Automated Forecast Generation

SAS-AM developed an automated works planning tool that ingests Maximo PM rule exports, joins them against the asset register and operational meter data, and generates a complete three-year forward maintenance forecast. The tool calculates next-due dates for every PM task on every applicable asset, then time-phases the resulting work orders across a rolling 36-month planning horizon.

The key technical elements included parsing Maximo’s PM rule structure to extract interval logic, trigger types, and task hierarchies; building a matching engine that correctly associates rules with assets based on classification, location, and applicability criteria; calculating next-due dates using last-completion data and current meter readings; and distributing forecast work orders across planning periods with resource levelling to avoid unrealistic peak loads.

The output is a structured dataset that feeds directly into Yarra Trams’ planning workflows — replacing the manual spreadsheet process entirely.

Results: From Weeks to Hours

The impact was immediate and measurable. The annual works planning cycle dropped from four to six weeks of manual effort to a matter of hours. Planners could regenerate forecasts on demand whenever PM rules changed or new assets entered service, rather than waiting for the next annual planning cycle. Forecast accuracy improved because the automated process eliminated transcription errors and ensured every applicable rule-asset combination was captured.

Perhaps most importantly, the tool gave Yarra Trams’ leadership team genuine visibility into the forward maintenance workload. For the first time, they could see the full three-year picture — identifying periods of peak demand, resource bottlenecks, and opportunities to optimise maintenance scheduling across the fleet.

The Broader Lesson

This engagement illustrates a pattern we see across asset-intensive organisations. The data needed for better planning often already exists within CMMS and EAM systems. The barrier isn’t missing data — it’s the absence of an analytical layer that connects disparate datasets and translates them into decision-ready outputs.

For Yarra Trams, the raw material was always there in Maximo. What was missing was the bridge between rule data and forward planning. Building that bridge didn’t require a new system or a major technology investment. It required understanding the data architecture, designing the right logic, and automating the translation from rules to forecasts.

If your organisation is spending weeks on planning processes that should take hours, the answer is probably already sitting in your CMMS. You just need the right approach to unlock it.