Maximo Health: How AI Scores Your Assets from 0 to 100 — And What to Do About It

Who this is for: Reliability engineers, maintenance managers, asset planners, and IT teams who need to understand what Maximo Health actually does, what data it needs to work, and whether it is worth deploying. Spoiler: it is the first suite application every organization should turn on.

Estimated read time: 10 minutes

The Question You Cannot Answer Today

Your VP of Operations walks into the Monday planning meeting. She drops a single question on the table:

"We have $5 million for capital replacement next year. Which of our 14,000 assets should we replace, which should we overhaul, and which ones are fine with current maintenance?"

You look at your reliability engineer. He opens a spreadsheet he built three years ago. It has 47 tabs. Half the formulas reference data that no longer exists. The last update was six months ago. He gives a number. Everyone in the room knows it is a guess.

This is the question Maximo Health was built to answer. Not with spreadsheets. Not with gut feel. With a score from 0 to 100 on every asset in your fleet, backed by data you already have in Manage, updated continuously, and fed into an optimization engine that tells you exactly where to put every dollar.

📊 Health Scores: One Number, Seven Inputs

A Maximo Health score is a single number from 0 to 100 that represents the overall condition of an asset. It is calculated from contributors — individual data points that each carry a weighted percentage of the final score.

Here are the seven contributor types and where each one comes from:

┌────────────────────────┬──────────────────────────┬───────────────────────────────────────┐
│ Contributor            │ Data Source              │ Example                               │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Age                    │ Manage (install date)    │ 15 yrs into a 20-yr expected life     │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Meter Readings         │ Manage (meters)          │ Runtime hours at 85% of expected life  │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Work Order History     │ Manage (work orders)     │ 12 emergency WOs in the past year     │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Replacement Frequency  │ Manage (inventory/WO)    │ Increasing parts replacement rate      │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Sensor Anomalies       │ Monitor (anomaly count)  │ 47 anomalies detected in 30 days      │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Inspection Results     │ Manage (inspections)     │ Last inspection score: Fair            │
├────────────────────────┼──────────────────────────┼───────────────────────────────────────┤
│ Custom Calculations    │ Custom formulas          │ Vibration trend slope > threshold      │
└────────────────────────┴──────────────────────────┴───────────────────────────────────────┘

Each contributor is assigned a weight. A typical configuration might look like: Age = 25%, Work Orders = 30%, Sensor Anomalies = 20%, Meters = 25%. You define the weights. Health calculates the combined score.

The result is a four-tier rating system:

┌───────────┬────────────────┬─────────┬──────────────────────────────────────┐
│ Score     │ Health Rating  │ Color   │ Suggested Action                     │
├───────────┼────────────────┼─────────┼──────────────────────────────────────┤
│ 80 - 100  │ Good           │ Green   │ Continue current maintenance          │
├───────────┼────────────────┼─────────┼──────────────────────────────────────┤
│ 60 - 79   │ Fair           │ Yellow  │ Monitor closely, increase PM freq    │
├───────────┼────────────────┼─────────┼──────────────────────────────────────┤
│ 40 - 59   │ Poor           │ Orange  │ Plan replacement or major overhaul   │
├───────────┼────────────────┼─────────┼──────────────────────────────────────┤
│ 0 - 39    │ Critical       │ Red     │ Immediate action required            │
└───────────┴────────────────┴─────────┴──────────────────────────────────────┘
Key insight: The score itself is not magic. The magic is that everyone in your organization — the VP, the planner, the technician — is looking at the same number. No more arguing about whether pump P-2047 "seems fine" or "looks like it's on its last legs." The score is 34. It is critical. The conversation shifts from opinion to action.

📉 Degradation Curves: Watching Health Decline Over Time

A single health score is a snapshot. A degradation curve is the movie.

Health tracks how every asset's score changes over time and plots it as a visual trend. These curves answer four questions that a point-in-time score cannot:

  1. How fast is this asset deteriorating? A pump that dropped from 80 to 60 in three months is a different problem than one that took two years to make the same decline.
  2. Is maintenance actually slowing degradation? If you performed a major overhaul at month 6 and the curve flattened, your PM program is working. If it kept dropping, you are spending money for nothing.
  3. When will health reach a critical threshold? Project the curve forward and you get a replacement date — not from gut feel, but from observed data.
  4. How does this asset compare to the fleet? Plot an individual asset against the average for its class and you immediately see outliers — the ones degrading faster or slower than their peers.

Degradation curves are the critical input for replacement planning and budget forecasting. Without them, you are planning capital budgets on calendar-based assumptions. With them, you are planning on observed condition data.

Hallway truth: "The degradation curve is what finally got our finance team to stop asking why we needed to replace a pump that was 'only' eight years old. The curve showed it was degrading twice as fast as the fleet average. They approved the PO the same day." — A reliability engineer at a mid-size manufacturer.

💰 Asset Investment Optimization (AIO): The $5 Million Question

AIO is Health's most powerful feature. It is the answer to the Monday meeting question we opened with.

Here is the problem AIO solves: Given a fixed budget, what is the optimal combination of maintenance, repair, and replacement decisions across your entire asset fleet?

That word "optimal" is doing serious work. AIO does not just rank assets by health score and replace from the bottom up. It uses optimization algorithms that simultaneously consider:

  • The cost of continued maintenance vs. replacement for each asset
  • Failure risk, health score, criticality, and remaining useful life
  • The compounding effect of deferred replacements
  • Budget constraints across a multi-year planning horizon

AIO Inputs

You feed AIO seven parameters:

┌────────────────────────┬───────────────────────────────────────────────────┐
│ Parameter              │ Description                                       │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Budget Ceiling         │ Maximum total investment for the planning period  │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Planning Horizon       │ Number of years to optimize over (1-30)          │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Asset Scope            │ Which assets or groups to include                 │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Replacement Costs      │ Per-asset or per-class replacement cost          │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Maintenance Costs      │ Annual maintenance cost per asset                │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Failure Costs          │ Cost of unplanned failure per asset              │
├────────────────────────┼───────────────────────────────────────────────────┤
│ Criticality Weights    │ How much to prioritize critical assets           │
└────────────────────────┴───────────────────────────────────────────────────┘

AIO Outputs

AIO returns four deliverables:

  1. Ranked action list — for every asset in scope, a recommendation: maintain, repair, or replace
  2. Year-by-year investment schedule — a capital plan that fits within your budget across the planning horizon
  3. Projected fleet health — what your overall fleet health score will look like under the recommended plan
  4. Cost-benefit analysis per asset — the financial justification for every recommendation
Key insight: AIO does not just tell you what to replace. It tells you the cost of NOT replacing. When you show finance a chart that says "deferring pump replacements by one year adds $800K in failure risk," the conversation about budget allocation changes completely.

Replacement Planning

Health also provides dedicated replacement planning views that sit alongside AIO:

  • Assets approaching end of life, sorted by urgency
  • Recommended replacement timelines based on degradation projections
  • Replacement cost projections rolled up by asset class, site, or organization
  • Impact analysis showing what happens if you defer replacement by 6, 12, or 24 months

These views turn AIO's optimization output into the format your capital planning team actually needs.

⚠️ Risk Matrix: Probability Meets Consequence

The risk matrix is one of Health's most immediately useful visualizations. It plots every asset on two axes:

  • X-axis: Probability of failure — derived from the health score and degradation trend
  • Y-axis: Consequence of failure — derived from the criticality score

This creates the standard risk quadrants:

                        CONSEQUENCE OF FAILURE
                    Low                    High
               ┌──────────────────┬──────────────────┐
    High       │                  │                  │
  Probability  │     MONITOR      │  ACT IMMEDIATELY │
    of         │                  │                  │
  Failure      ├──────────────────┼──────────────────┤
               │                  │                  │
    Low        │     ACCEPT       │  PLAN MITIGATION │
               │                  │                  │
               └──────────────────┴──────────────────┘

The power here is instant portfolio prioritization. You are not scrolling through a list of 14,000 assets trying to figure out where to focus. You are looking at a scatter plot where the top-right quadrant — high probability, high consequence — screams at you in red.

Every asset in that top-right quadrant is a ticking clock. Every asset in the bottom-left is something you can safely deprioritize. The matrix turns an overwhelming asset portfolio into a prioritized action plan in seconds.

🎯 Criticality Scoring: Not All Assets Are Equal

Health's criticality scoring defines the "consequence" axis of the risk matrix. It combines six factors into a single criticality rating:

Factor — What It Measures

Safety Impact — Injury or fatality risk if the asset fails

Environmental Impact — Spill, emission, or contamination risk

Production Impact — Revenue loss, throughput reduction, downtime cost

Regulatory Impact — Compliance violations, fines, shutdown orders

Financial Impact — Direct repair cost plus consequential damages

Redundancy — Is there a backup? Can operations continue without this asset?

A cooling water pump that serves a single production line with no backup and sits in an environmentally sensitive area will score very differently from a redundant utility pump with a standby unit. Both might have identical health scores. Criticality tells you which one keeps you up at night.

Hallway truth: "We had two compressors with the same health score — both around 45. One was in a redundant pair. The other was the single point of failure for our largest production line. Before Health, they showed up the same on every report. Now the risk matrix makes it obvious which one gets the capital dollars." — A plant manager at an oil and gas facility.

🔗 Integration Points: Where the Data Flows

Health does not live in isolation. It sits at the center of the MAS data pipeline:

┌──────────────┐        ┌──────────────┐        ┌──────────────┐
│              │        │              │        │              │
│   MANAGE     │───────>│   HEALTH     │───────>│   PREDICT    │
│              │        │              │        │              │
│ - Assets     │        │ - Scores     │        │ - Failure    │
│ - WO History │        │ - Curves     │        │   models     │
│ - Meters     │        │ - AIO        │        │ - RUL        │
│ - Inspections│        │ - Risk       │        │ - Anomaly    │
│ - Fail Codes │        │              │        │   detection  │
└──────────────┘        └──────────────┘        └──────────────┘
       ^                       ^
       │                       │
       │                ┌──────────────┐
       │                │              │
       └────────────────│   MONITOR    │
                        │              │
                        │ - Sensor data│
                        │ - Anomalies  │
                        │ - Alerts     │
                        └──────────────┘

Integration — Direction — What Flows

Manage to Health — Inbound — Asset records, work order history, meter readings, inspection results, failure codes

Monitor to Health — Inbound — Sensor metrics, anomaly counts, alert history

Health to Predict — Outbound — Health scores and contributors feed prediction models

Health to Manage — Outbound — Health scores visible on asset records, can trigger workflows

This is why Health is the first suite application you should deploy after Manage. It consumes data you already have. It produces scores that Predict needs downstream. It is the bridge between your operational data and your predictive analytics.

📋 Data Requirements: The Non-Negotiable Foundation

Here is where most Health deployments succeed or fail. The algorithm is only as good as the data you feed it.

┌───────────────────────┬──────────────────────┬────────────────────────────────────────┐
│ Data Element          │ Minimum Quality      │ Impact if Missing                      │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Install Dates         │ 90%+ populated       │ Age contributor will NOT function      │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Meter Readings        │ Weekly+ frequency    │ Meter contributor inaccurate           │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Work Order History    │ 2+ years             │ WO contributor unreliable              │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Failure Codes         │ Consistently recorded│ Cannot identify failure patterns        │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Asset Classification  │ Standardized hierarchy│ Cannot group for fleet analysis        │
├───────────────────────┼──────────────────────┼────────────────────────────────────────┤
│ Replacement Costs     │ Estimated per class  │ AIO cannot calculate ROI               │
└───────────────────────┴──────────────────────┴────────────────────────────────────────┘

Let's be blunt about each one.

Install dates at 90%+ population. If your asset records say "install date: null" on half your fleet, the age contributor is useless. Age is typically 20-30% of the overall score. Losing that contributor does not just reduce accuracy — it undermines confidence in the entire score. Fix this first.

Regular meter readings, weekly or better. A meter reading from six months ago tells Health nothing about current condition. If your technicians are not recording readings consistently, the meter contributor produces noise instead of signal.

Two or more years of work order history. Health needs enough history to establish patterns. How many emergency work orders per year? Is the frequency increasing? A brand new Manage deployment with six months of data will produce health scores, but they will not be trustworthy.

Consistent failure codes. If your work orders say "BROKE" in the failure code field — or worse, leave it blank — Health cannot distinguish between an electrical failure and a mechanical failure. Failure pattern analysis is dead on arrival.

Standardized asset classification. Fleet analysis requires grouping assets into classes. If your pumps are classified as "PUMP," "PMP," "Pump-Centrifugal," and "EQUIPMENT-ROTATING" across different sites, Health cannot compare them as a fleet.

Key insight: Data quality is not a Health problem. It is a Manage problem that Health exposes. If you deploy Health and the scores look wrong, do not blame Health. Look at your data. Every organization that has attempted to shortcut data quality has wasted their Health deployment. No exceptions.

⏱️ Pilot Effort: 42-84 Hours, 2-3 People

Here is the realistic effort to run a Health pilot:

Task — Effort — Prerequisites

Identify a pilot asset class (50-200 assets) with good data quality — 2-4 hours — Access to Manage data

Verify data quality: install dates, meter readings, WO history — 4-8 hours — Manage database access

Configure Health scoring group for pilot asset class — 4-8 hours — Health deployed, Manage connected

Define contributors and weights for the scoring group — 8-16 hours — Domain expertise from reliability team

Generate initial health scores and validate against known conditions — 8-16 hours — Scoring configured

Create degradation curves and analyze trends — 4-8 hours — Scores validated

Set up AIO scenario with realistic budget and replacement costs — 8-16 hours — Scores validated, cost data available

Present results to maintenance leadership for validation — 4-8 hours — All above complete

Total: 42-84 hours (1-2 weeks) with 2-3 people.

That is a reliability engineer who knows the assets, an IT resource who can configure Health, and a maintenance manager who validates that the scores match reality. If your pilot class has clean data, you can be presenting actionable health scores to leadership within two weeks.

This is the fastest time to value of any suite application. Monitor requires IoT sensor deployment. Predict requires Health scores plus training data. Visual Inspection requires camera hardware and model training. Health requires data you already have in Manage.

📊 Dashboards: Six Views That Change the Conversation

Health ships with pre-built dashboards that cover every angle:

  1. Fleet Overview — all assets plotted by health score with color coding. This is the "at-a-glance" view you put on the maintenance manager's screen.
  2. Health Distribution — histogram showing how many assets fall in each health range. If 40% of your fleet is in the red zone, you have a capital crisis. If 90% is green, you might be over-maintaining.
  3. Degradation View — time-series showing health score trends for individual assets or fleet averages. The "movie" that tells you where things are heading.
  4. Risk Matrix — the 2D probability-vs-consequence scatter plot. The view that prioritizes your entire portfolio in one screen.
  5. Investment Optimization — AIO results with budget scenarios. Show finance three scenarios: $3M, $5M, $7M. Let them see the health impact of each.
  6. Contributor Breakdown — detailed view of what is driving each asset's score. When someone asks "why is this pump at 34?", this is where you get the answer.

🆕 What is New in MAS 9

MAS 9 brings meaningful enhancements to Health:

Enhancement — What It Means for You

Enhanced scoring algorithms — More configurable contributor types, improved weighting options

Improved dashboards — Modernized UI with better drill-down capabilities

Better Monitor integration — Tighter coupling with sensor data as scoring contributors

Notebook-based custom scoring — Use Jupyter notebooks for custom scoring logic — no Java required

MAS 9.1 improvements — Additional visualization options, improved AIO performance

The Jupyter notebook integration is particularly significant. In MAS 8, custom scoring logic required Java development. In MAS 9, your data scientist can write a Python notebook that calculates a custom contributor — vibration trend analysis, dissolved gas ratios, whatever your domain requires — and plug it directly into the scoring engine.

🏭 Industry Use Cases

Utilities

  • Transformer health scoring based on dissolved gas analysis, load history, and age. A single transformer failure can cause millions in damage and weeks of outage. Health scoring identifies the transformers approaching failure before they get there.
  • Distribution pole replacement planning across thousands of assets. You have 50,000 wooden poles. You cannot inspect them all every year. Health scores prioritize which ones need physical inspection.
  • Substation equipment investment optimization. AIO tells you whether to replace the aging breaker or the degrading transformer when the budget only covers one.

Manufacturing

  • Production line equipment health based on vibration, temperature, and OEE data. A production line that drops from 85% OEE to 70% over six months shows a clear degradation curve — even if no individual alarm has fired.
  • CNC machine tool replacement planning. Spindle hours, tool change frequency, and dimensional accuracy trends feed the health score. When the curve projects a critical threshold at month 9, you order the replacement spindle at month 6.
  • Compressor and pump fleet health management. Across 200 pumps, Health identifies the 15 that need attention this quarter.

Oil and Gas

  • Pipeline segment health scoring based on corrosion data, pressure readings, and inspection results. Every segment gets a score. The ones in the red zone get inline inspection first.
  • Wellhead equipment degradation tracking. Remote wellheads are expensive to visit. Health scores prioritize which sites get a technician this month.
  • Rotating equipment fleet optimization. AIO balances overhaul vs. replacement across compressor fleets where individual units cost $2M+.

Transportation

  • Fleet vehicle health scoring from mileage, maintenance history, and age. A transit authority with 500 buses uses Health to plan annual fleet replacement cycles.
  • Rail switch and signal health management. Safety-critical infrastructure where failure consequence is measured in human risk, not just dollars.
  • Bridge and tunnel structural health integrated with Civil Infrastructure. Health scores combine inspection data with sensor readings for a holistic structural condition view.

The Decision Framework: Should You Deploy Health?

Answer these three questions:

  1. Do you have Manage deployed with 2+ years of operational data? If yes, you have enough history for meaningful scores.
  2. Are your install dates populated above 90%? If yes, your biggest contributor is ready.
  3. Can you identify a class of 50-200 assets with consistent meter readings and work order history? If yes, you have a pilot class.

If you answered yes to all three, deploy Health. It is 42-84 hours to a working pilot. It is the fastest path from "we think this asset is in bad shape" to "we know this asset is at 34 and here is why."

If you answered no to any of the three, fix your data first. Health will expose every data quality shortcut you have ever taken. Better to clean the data before deployment than to explain to leadership why half the fleet shows "Unknown" health scores.

Key Takeaways

  • Health scores (0-100) combine seven contributors — age, meter readings, work order history, replacement frequency, sensor anomalies, inspection results, and custom calculations — into a single condition indicator that everyone in your organization can understand
  • Degradation curves turn snapshots into trends — see how fast assets are declining, whether maintenance is working, and when health will reach critical thresholds
  • AIO answers the capital budget question — given a fixed budget and N assets, what is the optimal mix of maintain, repair, and replace? Show finance three budget scenarios and let the data make the case
  • The risk matrix instantly prioritizes your portfolio — high probability, high consequence assets jump off the screen. No more guessing where to focus
  • Data quality is the real prerequisite — 90%+ install dates, weekly meter readings, 2+ years of work order history, consistent failure codes, and standardized classification. Fix this first or do not bother deploying
  • 42-84 hours with 2-3 people — the fastest time to value of any suite application. You already have the data. Health just makes it visible

References

Series Navigation:

Previous: Part 9 — Maximo Monitor
Next: Part 11 — Maximo Predict

View the full MAS FEATURES series index

Part 10 of the "MAS FEATURES" series | Published by TheMaximoGuys

Maximo Health is where your asset data stops being a filing cabinet and starts being a decision engine. Every score, every curve, every AIO scenario turns raw operational history into capital planning intelligence. Deploy it first. Your $5 million question deserves a better answer than a spreadsheet with 47 tabs.