STRATEGIC MAINTENANCE SOLUTIONS, INC.
proven. reliable. trusted.
PREDICTIVE MAINTENANCE · CONDITION-BASED MONITORING
Sensors vs. Managed Monitoring
Can we just buy sensors, or do we need someone to watch the data?
TECHNICAL GUIDANCE PAPER
A practical guide for maintenance and reliability leaders deciding whether to run predictive maintenance (PdM) and condition-based monitoring (CBM) in-house with their own sensors and analysts, or as a managed service, from installation to work order.
Sensors collect data. Someone still has to decide what it means.
Up to 50%
Less unplanned downtimeSMS PdM deployments
24/7
Analyst-confirmed monitoringCertified SMS analysts
1,000+
Projects deliveredSince 1999
Five things to know before you buy
- 01The value is in the decision, not the data. Savings come from failures converted from unplanned to planned work, which requires a confirmed diagnosis and a work order.
- 02Timing decides everything. A fault is only worth detecting if the warning arrives early enough to plan the repair. Monitoring frequency must be set against the P-F interval of each failure mode.
- 03Generic thresholds create noise. Untuned alarms produce false positives that erode trust; alarms must be baselined to each asset and its criticality.
- 04Self-run programs carry hidden costs. Certified analysts, coverage for nights and vacations, IT/OT networking, cybersecurity, and integration are real line items, not extras.
- 05Most plants should start managed. Self-run, co-managed, and fully managed programs each fit a different team, but without certified analysts on staff, a managed pilot is the fastest route to proven results. The readiness check in Section 10 shows where you stand.
01 SENSORS ALONE
Sensors Alone: Data Without Decisions
A sensor-only program answers one question: “Is this machine’s vibration or temperature changing?” Industrial wireless sensors mount on motors, pumps, compressors, gearboxes, fans, conveyors, and robots, and stream condition data to a gateway and dashboard around the clock. What they don’t do is tell you what the change means, how serious it is, or what to do about it. That work still falls to your team.
Where sensor-only programs stall
What a wireless sensor actually sees
Wireless sensors trade data for battery life. Every measurement draws power, so sensors take periodic snapshots rather than continuous recordings, and battery life depends heavily on how often they sample and whether full waveforms are transmitted. Vendor battery claims are often quoted under conditions that aren’t clearly stated.
Specification matters. Industrial-grade triaxial sensors sample at 25 kHz or more, cover roughly 0.5 Hz to 10 kHz, and resolve thousands of spectral lines, enough for early bearing analysis. Lower-cost sensors may cap bandwidth near 1 kHz, missing the high-frequency bands where bearing defects first appear.
Slow equipment needs longer capture windows. On low-speed machines, fault impacts occur once per revolution and are spaced far apart; a short capture can miss them.
02 WHY TIMING MATTERS
Timing Is Everything: The P-F Interval
Detect early enough to plan, not just to watch.
Reliability engineers describe failure with the P-F curve. Point P is the earliest moment a developing defect can be detected; point F is functional failure, when the asset can no longer do its job. The time between them, the P-F interval, is the window in which a problem can still become planned work. Each detection method sits at a different point on the curve: high-frequency vibration and envelope analysis detect bearing defects earliest, while heat and audible noise appear only near the end.
Two practical rules follow
- 01The monitoring interval must fit inside the warning window. Published reliability guidance recommends checking at roughly one-half to one-third of the P-F interval, and one-quarter to one-fifth for critical assets, so there are at least two chances to catch the fault before failure.
- 02Detection is only half the window. Confirming the fault, ordering parts, and scheduling the repair consume the rest, which is why the speed of analyst confirmation and work-order creation matters as much as the sensor. This is the half of the window a managed service is built to protect: flagged data is reviewed continuously by analysts, so the diagnosis and work order follow the alert instead of waiting for someone to notice it.
Why generic thresholds misfire
Many sensor platforms ship with default alarm limits based on ISO 20816, the international standard for evaluating machine vibration, which replaced the ISO 10816 series. ISO 20816-3 sorts industrial machines above 15 kW into severity zones by broadband RMS velocity. These zones are a sound starting point, but they are machine-class averages:
| Machine group (rigid support) | Zone A/B | Zone B/C | Zone C/D | What the zones mean |
|---|---|---|---|---|
| Group 1: above 300 kW | 2.3 mm/s | 4.5 mm/s | 7.1 mm/s | A: new condition. B: acceptable for long-term operation. C: plan corrective action. D: damage likely. |
| Group 2: 15 to 300 kW | 1.4 mm/s | 2.8 mm/s | 4.5 mm/s | Flexibly mounted machines use higher limits; the group and support type must be set correctly. |
A single overall number cannot say which component is failing, and an early bearing defect can sit well inside Zone B while the damage progresses. ISO 20816-3 itself treats a step change of more than 25% of the Zone B/C boundary as significant even when the machine is still in an acceptable zone. Effective programs therefore combine ISO zones with asset-specific baselines, spectral band alarms, and envelope alarms, and they have a person confirm what the alarm means.
03 MANAGED PdM AND CBM
Managed Predictive Maintenance: From Signal to Work Order
Closing the loop.
Predictive maintenance delivered as a managed service answers a broader question: “Which assets are developing a fault, what is it, and what should we do about it this week?” It starts with the same sensors, then adds the people and processes that turn data into action: certified vibration analysts, alerting tuned to asset criticality, and automatic work orders in IBM Maximo or your CMMS.
The SMS approach
- Scope by criticality. An asset criticality review identifies the equipment whose failure costs the most, and the failure modes that matter on each, so sensors go where they pay back first.
- Install it right. SMS field technicians handle sensor placement and mounting, gateway and network commissioning, and panels and enclosures, with full as-built documentation.
- Match the sensor to the machine. Wireless sensors cover most assets; wired, continuous online systems with simultaneous multi-channel acquisition protect critical assets with short P-F intervals.
- Confirm before you alert. Edge machine learning flags anomalies, then an SMS analyst confirms the diagnosis with FFT spectrum, envelope, and trend analysis. Your team only sees what requires action.
- Close the loop in Maximo. A prioritized work order is created automatically with failure codes and recommended corrective actions. After the repair, fresh vibration data confirms the fix and sets a new healthy baseline.
Equipment covered
A managed program covers the rotating and production assets that drive most unplanned downtime, including legacy equipment, across single or multiple sites:
| Asset | Typical failure modes monitored |
|---|---|
| Motors | Bearing wear, imbalance, misalignment, looseness |
| Pumps | Bearing wear, cavitation, impeller wear, seal degradation |
| Compressors | Bearing faults, valve wear (rotary and reciprocating) |
| Gearboxes | Gear mesh, tooth wear, pitting, spalling |
| Fans and blowers | Imbalance, blade fouling, misalignment |
| Conveyors | Drive pulley bearings, tensioner and idler wear |
| Industrial robots | Joint reducer wear, servo motor health |
| Custom and legacy assets | Retrofit monitoring scoped to the failure modes that matter |
What you receive, every month
| Deliverable | What it means for your team |
|---|---|
| Analyst-confirmed findings | Each alert names the failing component, the failure mode, its severity, and the recommended fix. |
| Prioritized work orders | Created in Maximo or your CMMS, ranked by asset criticality and urgency. |
| Repair verification | Post-repair data confirms the fix worked and resets the baseline. |
| Program reviews | Regular reviews with cost avoidance, open findings, and system health, ready to share with leadership. |
| A team behind the data | Direct access to the SMS analysts who know your assets, for questions and troubleshooting during production. |
“Tony and his team have done a fantastic job keeping us on top of our other robotic health issues. We have been able to continuously plan, monitor, schedule and avoid these wear related failures.”
The analyst’s toolkit
No single technique catches every fault. SMS analysts combine deterministic vibration methods with AI-driven pattern recognition:
| Technique | What it reveals | Best for |
|---|---|---|
| FFT spectrum analysis | Converts the time signal into frequencies; each component has a signature (imbalance at 1× running speed, misalignment often at 2×, bearing defects at their own fault frequencies). | Identifying which component is failing on rotating equipment |
| Envelope / Hilbert demodulation | Isolates repetitive impacts riding on high-frequency carrier signals, exposing defects invisible in a standard spectrum. | Early bearing spalling, pitting, cracks; gear mesh faults |
| Time waveform analysis | Shows the raw signal shape and periodicity over time. | Slow-speed and reciprocating machines, impacting, cracked gear teeth |
| Wavelet transform | Time-frequency view of non-stationary signals. | Transient impulses and variable-speed machines such as robots |
| Trend and degradation forecasting | Correlates vibration with age, load cycles, and maintenance history. | Scheduling repairs weeks or months ahead |
One engagement, five phases
- 01
SCOPE
Criticality review
Asset criticality and the failure modes that matter.
- 02
DESIGN
Sensors & sampling
Sensor type, placement, sampling, integration.
- 03
BUILD
Configure
Sensors, gateways, connectors configured.
- 04
DEPLOY
Install & baseline
Install, commission, baseline, validate.
- 05
MANAGE
24/7 analysis
Analyst-confirmed alerts and Maximo work orders.
04 AT A GLANCE
Sensors Alone vs. Managed Monitoring at a Glance
| Dimension | Sensors Alone (Self-Run) | Managed Monitoring (SMS) |
|---|---|---|
| Core question | Is vibration changing? | What is failing, and what do we do about it? |
| Who reads the data | Your team, if it has trained analysts | Certified SMS vibration analysts |
| Analyst qualification | Varies; ISO 18436-2 Category II is the usual working minimum | Certified analysts, with senior review for complex faults |
| Alerts | Generic thresholds; high false-alarm risk | Analyst-confirmed, tuned to asset criticality |
| Diagnosis | Trend or alarm only | Specific failure mode with a recommended fix |
| Coverage | Business hours, unless you staff nights and holidays | 24/7 monitoring |
| Work orders | Entered manually, if anyone acts | Created automatically in Maximo or your CMMS |
| Sensor and gateway health | Your team tracks batteries and connectivity | Monitored as part of the service |
| Cybersecurity and networking | Your IT/OT team designs and secures it | Designed in from the start; reviewed with your IT team |
| Proof of results | Rarely measured | Cost avoidance and ROI reported at every review |
| Time to value | Months: hire or train, tune alarms, integrate | Weeks: scoped pilot on critical assets |
| In-house skills needed | Vibration analysis, IIoT networking, integration | Acting on confirmed recommendations |
| Recurring cost | Lower: hardware and software only | Higher: includes ongoing analyst service |
Rows in the middle of the table (qualification, coverage, sensor health, security, time to value) are where self-run programs most often under-budget, and exactly the work a managed service takes on.
06 THE BUSINESS CASE
Why It Matters: The Business Case
From alarms to avoided failures.
The value of condition monitoring comes from the failures you prevent, not the data you collect. The stakes are rising: Siemens’ True Cost of Downtime 2024 found that unplanned downtime now costs the world’s 500 largest companies about $1.4 trillion a year, roughly 11% of revenue, and that an hour of downtime in automotive plants costs about $2.3 million.[1]
$1.4T
Annual downtime cost, Fortune Global 500Siemens, 2024
11%
Of revenue lost to unplanned downtimeSiemens, 2024
$2.3M
Per hour of downtime in automotiveSiemens, 2024
8–12%
Savings of predictive over preventiveU.S. DOE / PNNL
What managed programs deliver in the field
| Business lever | Impact | Real-world / field evidence |
|---|---|---|
| Unplanned downtime | ↓ up to 50% | SMS PdM deployments deliver up to a 50% reduction in unplanned downtime; manufacturing sites typically see 40–60%. |
| Maintenance cost | ↓ 25–35% | Manufacturing deployments typically cut maintenance costs by 25–35%. |
| Maintenance efficiency | ↑ 30% | Condition-based scheduling raises maintenance efficiency about 30%. At Honda’s Lincoln, Alabama plant, it reduced technicians’ PM workload by eliminating time-based grease sampling, freeing them for higher-value work. |
| Repair timing | Planned, not unplanned | At Stellantis’ Sterling Heights plant, early alerts on critical robots let the team repair reducers during scheduled downtime, for a return several times the cost of the service. |
| Faster troubleshooting | Mechanical vs. electrical, remotely | At Cleveland-Cliffs, SMS analysts remotely ruled out a mechanical failure on a robot during production and pointed the team to the electrical fault that was the real cause. |
| Lost revenue avoided | $2.3M | At a power generation site, early bearing detection prevented $2.3M in lost generation revenue. |
What these results share: an SMS analyst confirmed each fault and a work order closed the loop. The sensors made them possible; the managed service made them happen.
Build the case on defensible numbers
Finance teams discount headline figures they can’t trace. The most conservative, well-sourced benchmarks are a good floor for your business case: the U.S. Department of Energy’s O&M Best Practices Guide puts predictive maintenance savings at 8–12% over a preventive program and 30–40% over reactive maintenance,[2] and Deloitte estimates that predictive maintenance raises equipment uptime and availability by 10–20% and cuts maintenance planning time by 20–50%.[3] Use these for the base case, then add your own failure history: every past failure on a monitored asset that could have been planned is a line item in the return.
“Right after the initial deployment, [the monitoring program] revealed several robots that were in various stages of failure that we were unaware of. This allowed our team to plan for the repairs and order the parts needed to keep our production floor running.”
07 TOTAL COST OF OWNERSHIP
The Real Cost of Running It Yourself
What the sensor quote leaves out.
Comparing a sensor price list with a managed-service subscription is not a like-for-like comparison. A self-run program needs everything the managed service includes; the cost simply sits in other budgets. Use the checklist below to price both options honestly.
| Cost element | Self-run: you own it | Managed: SMS owns it |
|---|---|---|
| Analyst staffing | At least one ISO 18436-2 Category II analyst; Category III for program design and complex faults. Certification requires classroom training, documented field experience, and an exam. | Included |
| Coverage and continuity | Backup for vacations, nights, and turnover; knowledge leaves when an analyst does. | Included, 24/7 |
| Baselining and alarm tuning | Setting and maintaining asset-specific baselines and alarm bands as operating conditions change. | Included |
| Networking and cybersecurity | Gateway placement, wireless coverage, segmentation, firewall rules, and security reviews with IT. | Designed and documented by SMS |
| CMMS / EAM integration | Building and maintaining the Maximo or CMMS interface, failure codes, and work-order templates. | Included (IBM Platinum Business Partner) |
| Fleet upkeep | Battery replacement, firmware updates, failed sensors, and gateway outages. | Monitored; SMS field service |
| Software and data | Platform licenses, data storage, and user management. | Included in subscription |
| Reporting | Building the cost-avoidance case for leadership. | ROI reported at every review |
The talent gap is real
Analyst skills are scarce and getting scarcer. Deloitte and The Manufacturing Institute project that U.S. manufacturers may need as many as 3.8 million new workers between 2024 and 2033, and that about 1.9 million of those jobs could go unfilled without significant change.[4] A program built around one or two experienced analysts is exposed to that market.
Price these line items honestly and the gap between a sensor quote and a managed subscription narrows quickly, before counting the failures an understaffed program misses. For many plants, that makes the managed path the lower-risk choice, not just the faster one.
08 SLOW- AND VARIABLE-SPEED ASSETS
Special Case: Slow- and Variable-Speed Assets
Where generic sensors struggle most.
Most off-the-shelf vibration programs assume a machine runs at a steady, moderate speed. Many critical assets don’t: low-speed gearboxes, conveyors, mixers, and rolls turn slowly, while industrial robots and indexing machines change speed and direction many times per cycle under varying loads.
On slow machines, a bearing or gear defect produces an impact only once per revolution, so impacts are spaced far apart and a short capture window can miss them entirely. On variable-speed machines, fault frequencies move with every speed change, smearing a standard spectrum. Research on robot joint reducers illustrates the point: vibration analysis is the most widely used and mature diagnostic method, but variable-speed operation produces non-stationary signals that fixed-speed spectra handle poorly.[10]
How a managed program handles it
These assets are where analyst expertise matters most, and where self-run programs most often produce noise instead of findings. A managed program addresses them by design:
“[The program] helped us to predict these failures, plan and schedule for the replacements, and create a better robotic health maintenance program.”
09 BUILD OR BUY
Honest Trade-Offs: Build, Buy, or Share the Work?
The case for running it yourself
Organizations with an established reliability team, certified vibration analysts on staff, and IT support for industrial networks can run a sensor program in-house. Recurring cost is lower, the expertise stays inside the company, and the team controls thresholds and priorities directly.
The case for a managed service
Most maintenance teams weren’t built for data analysis, and hiring or training vibration analysts takes time most plants don’t have. The common failure mode is a pilot that produces alerts nobody acts on until the program loses credibility and stalls. A managed service removes that risk: analysts are in place from day one, alerts arrive already confirmed, and results are reported in terms leadership cares about.
The middle path: co-managed
Many plants fit neither extreme. Your team owns routine data review and first-line response, while SMS analysts handle 24/7 coverage, complex diagnoses, and escalations, and mentor your staff toward certification. It builds internal capability without leaving critical assets uncovered.
| Self-run | Co-managed | Fully managed | |
|---|---|---|---|
| Best when | Certified analysts on staff and a mature reliability program | Some in-house skill; want to grow it with expert backup | No vibration program yet, or analysts are stretched thin |
| Your team does | Everything | Routine review, first response, repairs | Repairs on confirmed recommendations |
| SMS does | Installation and integration (optional) | After-hours coverage, complex diagnosis, coaching | The full loop, scope to verified repair |
| Cost profile | Lowest recurring; highest internal labor | Moderate | Highest recurring; lowest internal labor |
10 READINESS CHECK
Readiness Self-Assessment
Which path fits your plant today?
Answer each question honestly for the assets you plan to monitor. Check every statement that is true today.
Your score
0/ 8
0–2
Start with a managed pilot
Begin on your most critical assets, and build internal capability as results come in.
3–5
Co-managed, or managed to start
Keep what your team does well; let SMS cover analysis depth and after-hours coverage.
6–8
Self-run is viable
Consider SMS for installation, integration, and periodic program audits.
Every “no” is work a managed service takes on from day one.
11 START RIGHT
Designing a Pilot That Proves Value
Small scope, clear success criteria.
A pilot should answer one question for leadership: did monitoring convert failures that would have been unplanned into planned work, at a cost that makes sense? Agree on how that will be measured before the first sensor is installed. SMS runs pilots as managed engagements, so your team spends the pilot judging results rather than building infrastructure.
- 01
WEEKS 0–2
Scope
Criticality review; select 10–30 assets with high failure cost and vibration-detectable failure modes; pull CMMS failure history; agree on KPIs.
- 02
WEEKS 2–4
Design and build
Sensor selection and placement; gateway and network plan reviewed with IT; Maximo or CMMS integration configured.
- 03
WEEKS 4–6
Deploy and baseline
Install and commission; capture healthy baselines; set alarm bands by asset and criticality.
- 04
WEEKS 6–12+
Manage and review
Analyst-confirmed alerts and work orders; monthly reviews; a results readout with cost avoidance and a scale-up plan.
Measure what matters
| KPI | What it shows | Target |
|---|---|---|
| Confirmed findings | Number of analyst-confirmed faults, by failure mode and severity | Set at kickoff |
| Alert precision | Share of alerts that led to a confirmed, actionable finding (the inverse of alert fatigue) | Set at kickoff |
| Detection lead time | Days between first alert and the planned repair date | Inside the P-F window |
| Work-order closure | Recommended work orders completed, and how many were verified fixed by follow-up data | Set at kickoff |
| Failures avoided and cost avoidance | Events converted from unplanned to planned, valued with the ROI worksheet in Section 07 | Reported at every review |
| System health | Sensor and gateway uptime; data completeness | Set at kickoff |
Settle security and data questions up front
Every connected sensor and gateway extends the plant’s attack surface, so bring IT/OT security into scoping, not commissioning. The ISA/IEC 62443 series is the international reference for securing industrial automation and control systems, and its zone-and-conduit model is a practical way to frame the conversation.[11] Questions to answer in week one:
- Network path. Will sensor data travel on a segmented OT network, a dedicated cellular gateway, or the plant network? Which firewall rules are needed?
- Encryption and authentication. Is data encrypted from sensor to gateway to platform, and are devices and users authenticated?
- Data ownership. Who owns the raw data and the diagnoses, can it be exported, and how long is it retained?
- Access. Which plant, corporate, and SMS users can see which assets, and how is access reviewed?
What a managed pilot includes
| Included | Detail |
|---|---|
| Criticality review | Asset selection and failure-mode review with your reliability team |
| Installation | Sensors, gateways, and enclosures installed and commissioned by SMS technicians |
| 24/7 monitoring | Analyst-confirmed alerts with a specific diagnosis and recommended fix |
| Work orders | Prioritized work orders in IBM Maximo or your CMMS |
| Results readout | Cost avoidance, findings, and a plan to scale, ready for leadership |
12 THE BOTTOM LINE
Data is cheap. Decisions prevent failures.
Sensors are the starting point, not the result. Downtime falls only when someone confirms the fault, sets the priority, and gets a work order to a technician in time. If your team can’t do that today, buy the outcome: start with a managed pilot on your most critical assets and let the results make the case.
PARTNER WITH SMS
As an IBM Platinum Business Partner and an authorized Siemens Senseye implementer, with 25+ years of engineering and maintenance expertise, 1,000+ projects delivered, and more than 50,000 assets monitored across six continents, Strategic Maintenance Solutions helps organizations move from raw sensor data to a fully managed predictive maintenance program, from the first pilot to enterprise scale.
Start with a criticality review: talk to an SMS reliability engineer about a managed pilot. Contact sales@sms-inc.net or +1 207-854-9400.
A APPENDIX
Glossary and References
Glossary
| Term | Meaning |
|---|---|
| CBM | Condition-based maintenance or monitoring: performing maintenance based on the measured condition of an asset rather than a fixed schedule. |
| CMMS / EAM | Computerized maintenance management system / enterprise asset management system, such as IBM Maximo, where work orders are planned and tracked. |
| Envelope analysis | A demodulation technique that isolates repetitive impacts from bearing or gear defects that are hidden in a standard spectrum. |
| FFT | Fast Fourier transform: converts a vibration signal over time into a spectrum of frequencies. |
| ISO 18436-2 | International standard for qualifying vibration analysts, in four categories (I to IV) of increasing skill and responsibility. |
| ISO 20816 | International standard series for measuring and evaluating machine vibration, with severity zones A to D. Replaces ISO 10816. |
| ISA/IEC 62443 | International standard series for cybersecurity of industrial automation and control systems. |
| PdM | Predictive maintenance: using condition data and analysis to forecast failures and schedule repairs before they occur. |
| P-F interval | Time between the first detectable sign of a defect (P) and functional failure (F). |
References
- [1] Siemens (Senseye Predictive Maintenance). The True Cost of Downtime 2024.
- [2] U.S. Department of Energy, FEMP / Pacific Northwest National Laboratory. Operations & Maintenance Best Practices Guide, Release 3.0.
- [3] Deloitte Insights. Making maintenance smarter: Predictive maintenance and the digital supply network.
- [4] Deloitte and The Manufacturing Institute (2024). Taking charge: Manufacturers support growth with active workforce strategies.
- [5] ISO 18436-2:2014. Condition monitoring and diagnostics of machines: Requirements for qualification and assessment of personnel, Part 2: Vibration condition monitoring and diagnostics.
- [6] ISO 20816-1:2016 and ISO 20816-3:2022. Mechanical vibration: Measurement and evaluation of machine vibration.
- [7] Reliable (ReliaMag). Early Failure Detection: What Plants Get Wrong About the P-F Curve.
- [8] Spence, E. and Pucillo, J. What Should I Know About Wireless Sensors for Asset Condition Monitoring? Reliable Plant.
- [9] Murata. Why early bearing faults are difficult to detect wirelessly.
- [10] Lv, Z. et al. (2024). A Review and Prospect of Fault Diagnosis and Condition Monitoring Methods for Rotary Vector Reducer. CCESEM 2024, DOI 10.2991/978-94-6463-606-2_51.
- [11] ISA Global Cybersecurity Alliance. OT Cybersecurity for Industrial Sensors? The Why and the How.
- [12] Strategic Maintenance Solutions. Predictive Maintenance Field Services & Solutions (2026); customer testimonials (2025). www.sms-inc.net
Industry figures are cited from the published sources above. SMS field results reflect specific client deployments; results vary by asset, failure mode, and operating conditions. IBM Maximo is a trademark of IBM. Siemens Senseye is a trademark of Siemens AG.
FREE ASSESSMENT
Request a PdM Assessment
Start where the value is. An SMS reliability engineer will review your critical assets with you and show where condition monitoring pays back first, whether you run it yourself, share the work, or hand it to us.
- Asset criticality review of the equipment whose failure costs the most
- Failure modes and monitoring fit for each asset in scope
- A recommended path: self-run, co-managed, or a managed pilot