Wireless Vibration Sensor Retrofit: Procurement and Acceptance Guide

When retrofitting wireless vibration sensors to existing factory equipment, simply selecting candidates by comparison charts (such as measurement items or battery life) and placing an order may still result in a system that cannot pass operations acceptance. You’ll need to consider: will it mount securely to the bearing? Will the radio signal reach reliably under actual running conditions? Is missing data (dropouts) visible? Who exactly will handle battery replacements? Incorporate these factors into your estimation specifications (RFP) and acceptance test items. This article is a practical guide for factory purchasing, maintenance, and OT professionals in Thailand who wish to implement wireless vibration monitoring on existing motors or pumps. For a discussion of detectable and undetectable faults via vibration, and calculation of ROI, refer to our article on equipment diagnosis with vibration sensors. Here, we focus solely on “what to purchase and accept after you decide to proceed.”

You Can’t Order “Sensor Only” When Retrofitting Wireless Vibration Sensors

To establish one measurement point, you need more than just the sensor itself: consider the mounting surface on the machine, the sensor unit, fixtures, battery, wireless network, gateway, data storage, time synchronization, interface, maintenance notifications, and users. Published “wireless range” specs don’t guarantee radio performance in real production environments with metal equipment, piping, fences, vehicles, etc. Likewise, spec-sheet battery life can’t be applied without considering the effects of measurement cycles, waveform acquisition, retransmission, temperature, communications path, and alert settings. So, compare not “the per-unit price,” but the system as a whole —“the complete set that reliably delivers actionable data at the specified measurement point to your maintenance team.”

For example, ifm’s wireless VWV system sends periodic overall vibration velocity and temperature to a gateway. SKF’s Enlight Collect IMx-1 measures acceleration, enveloping, velocity, and temperature—a setup comprising battery-powered sensors, gateways, apps, and monitoring software. The same “wireless vibration” label hides major differences in data and system requirements. When comparing, do NOT score blanket terms like “wireless” or “AI-compatible”; instead, confirm, by model and software version, what values needed for target fault detection you can acquire, intervals, time stamps, quality info, re-acquisition capability, and output/format.

Divide your purchase specs into at least the following five categories: (1) target machines and measurement point scope, (2) mounting method and safety conditions on-site, (3) performance for measurement, communication, and dropouts, (4) operations responsibility, including battery/software/network costs, and (5) test evidence—Factory Acceptance Test (FAT) and Site Acceptance Test (SAT)—for pass/fail criteria. Without this breakdown, you risk suppliers claiming “the sensor is fine,” maintenance saying “there’s no data,” and IT saying “it’s out of network range.”

Start by Choosing Measurement Points: Where on Existing Motors to Mount

Do not choose the mounting location as just any free surface on the exterior. Base your choice on failure hypotheses: locate near bearing housings or rigid structural members transmitting loads, and note measurement orientation, drive/non-drive sides, and the directions (horizontal, vertical, axial) in your measurement point register. As a general mounting example, ifm’s guidelines for its wired IO-Link VVB3 recommend placement close to bearing seats, list threaded, magnetic, or adhesive adapters, and warn that mounting accessories affect measurements. SKF’s IMx-1 manual specifies axis sensitivity must be perpendicular to mounting surfaces, with correct site prep, ample space for tools, and access for maintenance. So confirm not just “how many sensor axes,” but “can the required direction be measured at each mounting location.”

For retrofits, you may face old paint, curved/thin covers, heat sources, washdown water, or vibrating brackets. Don’t assume the magnetic fixtures you use for provisional testing are suitable for permanent installations. High-frequency transmission depends significantly on mounting method—verify the measurable frequency range for each method by checking both manufacturer literature and on-site trial measurements. SKF’s sensor catalog explains frequency response per mounting method, offering adhesive mounting if proper hole-tapping isn’t feasible. Threaded mounting isn’t always practical; adhesives depend on surface and environment. Separate fixtures and mounting work from sensor procurement in your acceptance checklist.

The measurement point register should record equipment ID, location of rotating elements and bearings, photos, direction, sensor/fixture part numbers, installation date, responsible personnel, surface treatment, tightening or adhesive method, protective covers nearby, and maintenance access for replacement. After installation, compare values at the same direction and load using portable meters or reference points—investigate any notable discrepancies. However, never compare numbers alone if instruments or filtering settings differ. Your acceptance checklist should include units, frequency bands, integration settings, averaging method, rotation speed, and load.

Installation itself is subject to conditions: if you must approach rotating sections or hot surfaces during operation, follow site-approved isolation and lockout procedures. “Wireless = no need to stop” does NOT hold. Removing covers, hazardous zone certifications, washdown or food-safety requirements, and OEM warranties must be separately checked. These sensors are for monitoring only—not for direct safety/trip circuit substitution.

Data Specs to Check When Selecting Wireless Vibration Sensors

In the RFP, ask not only what measurements are provided, but “which signal, when, using what process, was this value generated?” Monitoring overall vibration velocity may suffice for trends, but for acceleration waveforms or enveloped data, required frequency range, sampling, storage, battery use, and bandwidth change. NI notes the Nyquist theorem: sample at more than twice the highest component frequency and use anti-aliasing filters. This is only a theoretical minimum—whether the actual sampled waveform meets diagnostic aims isn’t guaranteed by this number alone. The RFP should specify: target faults, required frequency range, what waveform/spectrum/feature data can actually be stored or output, and filter properties.

“Data arrives every minute” needs clarification. Is the sensor measuring every minute? Is the gateway sending the prior measurement to the screen every minute? Or just evaluating for alarms every minute? Separate acquisition time and receive time; check acquisition interval, calculation window, regular transmission interval, extra sampling for alarms, on-demand measurement, push-after-recovery, and storage on device. SKF documents explain that, to save power, terminal units sleep between configured acquisition intervals and just send aggregate data—don’t assume short-duration anomalies are always caught.

This table does NOT indicate absolute performance standards. For example, don’t set a “permissible dropout rate” as a universal value upfront—instead, decide this per asset: how quickly faults escalate, when maintenance checks in, if a backup round exists. Never leave permissible values for each asset blank in the contract. On the acceptance form, record equipment and measurement-point IDs, operating and stopped conditions, test start and end times, the approver, and the software version. A simple pass tick leaves no evidence of how a communications failure was reproduced or how long recovery took. For each failed item, log whether temporary operation is allowed, the retest date, an alternative manual inspection, and the owner of corrective action. A measurement that works in a vendor lab is different from a maintenance record usable amid the plant’s metal obstructions and actual loads. Accept the system using evidence that can be reproduced on site under the same conditions.

Local Wireless Survey: Don’t Trust Catalog Range—Test from Each Measurement Point

Radio environments in factories change with metal walls, pipes, switchboards, forklifts, shutters, layout shifts, and new installation work. Mesh wireless does not guarantee unconditional, stable communication for every point. Emerson’s wireless deployment guides include line of sight, height, equipment density, device/gateway arrangement for on-site surveys. SKF’s IMx-1 manual also stresses clear lines of sight between the gateway and sensors; when placing a gateway inside a metal enclosure, an external antenna is called for. Use actual or equivalent hardware to test radio conditions at all candidate points before choosing a system.

Mapping should include measurement point numbers, installed heights, metal obstructions, candidate gateway positions, power outlets, external network access points, and plans for future expansions. Don’t restrict tests to quiet holidays; include typical production, vehicle movement times, shutter operations, and line changes. Don’t judge solely by signal strength—record, per terminal, packet reception ratios, latency, retransmission, path changes, periods of no data, and gateway recovery after downtime. Expanding the number of devices will change the network—don’t overgeneralize PoC results with a handful of working points to the entire facility.

Certain wireless types may require compliance with local radio law, band, certification, power, and SIM contract. For Thailand, confirm with the supplier if the model number and firmware revision are eligible for local use, and ask for documented proof. This article does not guarantee any given product’s suitability in Thailand. Network connections must be approved by OT and IT leaders—confirm monitoring (not control) system write isolation. Refer to NIST SP 800-82 Rev.3 for OT security, reliability, and safety requirements. Record the blueprint of which ports, authentication, update paths, logs, and backups are necessary on the gateway.

Don’t Treat Dropouts and Time Stamps as “Zero Value”

It’s crucial in wireless monitoring to distinguish between “low vibration” and “no data received.” Never treat “no data measured,” “measured but failed to send,” “reached the gateway but not the server,” “out-of-sync times,” or “late uploads” as the same blank/zero. Display data values and quality status in different fields in both interface and API. If data analytic models “learn” that “normal during dropouts,” alarm significance is lost.

Sample acceptance includes, for each record: equipment ID, measurement point ID, sensor ID, acquisition timestamp, receive timestamp, unit, value, quality code, configuration version. Where these requirements can’t be met, agree in advance on alternatives and traceability. FAT should include planned stoppages (cutting sensor/gateway communications) to verify the display does not show old values as current, that dropout start time is logged, and post-recovery replay and deduplication work as specified. SAT should repeat these tests during live plant operation, including onsite network and notification path.

Time is not just display format. Correlate vibration data with equipment stoppages, speed changes, and maintenance activity—document device times, gateway times, server times, time zones, and synchronization sources. In Thai factories, screens may use the local time (ICT) and servers UTC. Document timezone handling and daylight saving rules, making sure the same event gets recorded as the same time across all systems. On replayed data, retain original acquisition time; don’t overwrite with receive time—this aids traceability.

Don’t Compare Battery Life by “Years”—Estimate as an Operational Task

Product-page battery life is strictly conditional. For instance, ifm’s VWV001 page specifies the battery isn’t replaceable; SKF Enlight Collect data shows “up to 4 years” depends on configuration. The manufacturer currently labels VWV001 a phase-out article. For a new purchase, confirm the current replacement part number and do not compare “years” across products under different settings. Side-by-side comparisons should state product name/part no./firmware, measurement and transmit intervals, extra waveform captures, temperature, comms path, replaceability, and require the supplier’s technical justification under identical conditions.

Management must go beyond just the battery. Factor in: the person who does replacement, need for shutdown/high-access/scaffolding, stock/resupply lead times, ID reassignment and measurement-point mapping after swap, used battery disposal, handling of dropouts during change, value baseline confirmation after change. If swapping batteries is impossible, price full device replacement and recovery. Installing where you can’t do replacements raises O&M costs, no matter how cheap the unit price. Battery level alerts must be set to provide adequate advance notice for maintenance’s lead times.

Clarify “estimated life” versus “warranty” in contracts. “Estimated life” is a figure with conditional assumptions or a manufacturer claim; “warranty” specifies fixes and responsibility on failure. If early depletion occurs, write down how to isolate root cause (terminal defect, excessive retransmission, changed setting, temperature, battery lot, wireless environment) and which logs to check. Spare parts plans should cover not just numbers, but storage, expiry, and handover of config/settings.

Don’t Accept PoC Success Based Only on “Getting a Graph”

Trial wireless vibration monitoring on a small number of existing motor points is NOT about “catching real failures” in a short period. Even when no faults arise during the test, you can verify reproducibility of installation, sustaining communications, interpretation of data, and notification responses. Before starting PoC, register target equipment, points, running conditions, expected faults, and current inspection methods; agree on evaluation period and success criteria. Do NOT cause intentional faults on site—rather, plan safe simulated events, such as comms interruptions, terminal swaps, config changes, and notification drills.

Don’t lock the “baseline” into a fixed average during the test. Vibration changes with speed, load, process mode, product, phase (startup, warm-up, pre- and post-maintenance). Display “baseline not established” for conditions with too few samples; do NOT pretend thresholds are set automatically. Log who created the baseline, coverage period, excluded abnormal operations, data quality, and approval date. On any alert, ensure maintenance staff check the real unit, then log cause and action path—that linkage, not more sensors, is what makes maintenance effective.

Evaluation sheets must capture planned/actual measurement count, valid data, dropouts, latency, false alarms, confirmed anomalies, rework on installations, battery reading variances, and work time. Do NOT cite absence of failures as “100% predictability” or “demonstrated fault reduction” after a short PoC. Real effect requires long-term tracking of actual downtimes and maintenance action, with conditions clarified. Our article on aging equipment IoT covers frameworks for continuing/renewing/retiring existing assets. This PoC phase is just to decide whether wireless vibration monitoring kits are suitable.

Clauses to Include in Your Wireless Vibration Sensor RFP

Begin the RFP by stating project objectives and what’s out of scope. The objective: monitor equipment mechanical status for maintenance tracking and action. Out of scope: automatic safety trips, protective relays, PLC reprogramming, or full automation of any possible fault prediction. Next, provide an asset list, candidate measurement points, running hours, temperature/washdown/hazards, available power/networking/windows for mounting, and local photos/drawings. If any fact is undetermined, mark it so, and set terms for revising pricing after site survey. Pressuring suppliers to squeeze unknowns under a “standard installation” only invites future upcharges.

Require that proposals break down sensors, mountings, gateways, antennas, power, comms, software/licenses, installation, testing, training, support, and spares line by line. Confirm part/model, revision, lead time, warranty, license renewal/end, data ownership/export, and migration if supplier withdraws. Separately list costs that are fixed per device, and those that scale with more points—don’t compare purely by lowest sensor price.

Technical conformance sheets should have more than “compliant/not compliant.” For each requirement, cite supporting manufacturer docs, proposal evidence, FAT test, SAT test, exceptions/alternatives, and responsible parties. When comparing different wireless types, judge using identical measurement points, dropout definitions, and running conditions. Do not compare “years” between proposals that extend sensor life by cutting interval versus those that prioritize frequent waveform acquisition. Fix monitoring frequency for critical equipment first.

What to Accept in FAT (Factory) & SAT (Site) Testing

FAT verifies, pre-shipment, that functionality is in place for the specified config and software. SAT checks, after field installation, the full workflow in actual environment (equipment, RF, network, user). A system may pass FAT signal checks and still fail on-site due to unexpected metal obstacles. A chart appearing in SAT doesn’t guarantee robust data persistence or correct handling of device swap-outs and downtime. Assign distinct goals to each phase.

Numeric checks must be recorded per the agreed RFP for each measurement point: comms reception rate, max continuous data gap, alarm lag, time deviation, data retention term, battery alert lead days — “criteria,” “method,” “test period,” and “exceptions” must be documented. No universal “threshold” values are assumed—acceptable gaps run from one hour to a day, depending on asset risk and local response time.

Decide up front how to handle failures. Who pays to move a gateway or re-test after antenna position changes? Does moving a point mean re-baselining? Is there a retest deadline? SAT signatures must include not just procurement, but maintenance, OT/IT, and, if relevant, safety—avoid equating “installation completed” with “system accepted.”

Operations Handover and Change Management After Go-Live

Wireless vibration monitoring is not a “one and done” task at installation. Maintenance owns the measurement points, thresholds, inspection, and replacement duties; OT/IT manages networks, accounts, time, backups; procurement manages warranty, spares, contract renewals. The supplier provides troubleshooting logs and procedures. If you don’t clarify first-line troubleshooting, communication dropouts will paralyze all sides—sensor, wireless, and server staff waiting on each other.

Actions like bearing change, motor replacement, moving sensors, gateway swap, firmware upgrade, or interval change disrupt data continuity. Log the event date, settings before/after, affected points, need for reconfirmation, and whether to transfer baselines. Even if the trend line looks seamless, any shift in mounting or data conditions breaks the historical series. When replacing a terminal, keep sensor and equipment ID histories separate, don’t erase pre-change data.

Monitoring infrastructure itself requires regular checks: receive halt, time sync loss, battery level drop, storage full, notification target turnover, cloud contract lapses—all risk missing real faults. Monthly checks must cover not just unreconciled alerts, but each point’s data yield, dropouts, battery/sensor status, gateway uptime, and backup results. Monitor metrics using the same definitions as SAT, and clearly document reasons when changing any KPI definition.

Easy-to-Miss Handover Items in Quotation Comparisons

If the installation quote says only “full system setup,” insist on detail: measurement point registry, install photos, wireless route map, gateway config backups, terminal settings list, account ledger, firmware version, data item definitions, API specs, time-sync configs, dropout code lists, notification contact list, battery replacement procedure, spares log, and original FAT/SAT records. “You can see it on the supplier’s cloud” doesn’t mean you have what’s needed for change tracking or recovery. Do not simply include confidential data in unprotected docs; designate credential storage and administrator handover.

The maintenance contract must distinguish hotline contact for sensor, wireless, or software failure, response times, on-site visit terms, and spare supply periods. If you can swap a device, but firmware or gateway versions are incompatible, recovery will stall—ensure migration and compatibility for major changes. Before contract award, test if you can export acquired waveforms/features in industry-standard formats if/when you stop using the vendor’s software. Running extraction tests for future system migration is more convincing than any demo.

Separate scope finalized post-survey (ex: antenna, power work) from per-unit prices settled at start (ex: sensor, standard mount bracket), settling extras by as-built drawings and quantities—as opposed to prioritizing a low initial quote and then later cutting scope on-site. For exceptions, record—per measurement point—the reason, resulting risk, backup inspection path, and approver. Avoid vague postponement of acceptance or operational go-live; clarify cut-offs for liability after install.

FAQs: Points to Confirm Before Wireless Vibration Sensor Deployment

Are wireless vibration sensors really easier to retrofit than wired types?

They can reduce cabling, but require mounting surface prep, wireless survey, gateway power, network approval, battery change process, and dropout tracking. Judge not just total install time but maintenance rounds and replacement actions. If you require 24/7 waveform capture or ultra-low-latency triggers, battery-powered, periodic sensors may be unsuitable.

Is it okay to start monitoring vibration on an existing motor using a magnetic mount?

It works for provisional/parallel comparison, but permanent mounting method must fit required frequency band, surface shape, temperature, washdown, and loss prevention. Confirm manufacturer’s recommended mounting and compare with hand-held or portable measurements at actual location. Follow site safety for exposed/moving parts or shield covers.

How do I handle data loss/dropouts with wireless vibration sensors?

Never fill missing values with zero! Record each as “not measured,” “transmit failed,” or “post-receipt transfer failed,” and distinguish acquisition vs. receipt time, dropout start/recovery, retransmission, and deduplication—test all these. Assign a responsible person to check/replace sensors or chase comms issues if downtime exceeds allowed time.

Should both FAT and SAT be required in vibration monitoring RFPs?

Yes. Factory pre-shipment config/functionality and onsite mounting/RF/network/user acceptance are separate evidence. Even for small projects, keep their purposes and signoff authorities separate so you can define what “it worked” actually means. Run simulated comms cut (FAT) and real-world operator alert checks (SAT).

How should I compare battery life estimates?

Compare NOT by max/average years, but under aligned measurement intervals, waveform frequency, communications path, temperature, and battery replacement method. Ask the supplier for a conditional technical basis and how to isolate premature battery drain causes.

Conclusion: Prepare Specifications and Test Sheets First

For retrofitting wireless vibration sensors, first decide measurement points, validate mounting, survey on-site wireless, separate values/dropouts, clarify battery replacement roles, and document evidence by FAT and SAT. Price comes later. Compile one-pager specs for a key asset with registry, needed data, permissible loss, replacement criteria, and test items—you’ll spot proposal gaps and pending issues.

At TOMAS TECH, we can help Thai factories structure requirements for wireless vibration monitoring on existing assets—including measurement points, system integration, and local acceptance. Even if your equipment or vendor is undecided, with photos and asset info you can contact us for consultation.

References

• SKF Enlight Collect IMx-1 System User Manual, Revision L: Mounting orientation/surfaces, gateway layout, and system architecture.
• SKF vibration sensor catalogue: mounting requirements: Effects of mounting methods on measurable frequency.
• SKF Enlight Collect IMx-1 data sheet: Intermittent sampling and configuration-dependent battery life.
• ifm wireless VWV overview: Transmitted values and gateway setup.
• ifm VVB3 installation guidelines: Bearing-seat placement, mounting options, and accessory effects.
• ifm VWV001 product page: Battery replacement specifications and phase-out article status.
• Emerson Wireless Projects Service: Site survey key points.
• NI: Acquiring an Analog Signal: Sample rate and anti-aliasing theory.
• NIST SP 800-82 Rev.3: OT security incorporating performance, reliability, and safety.