vaayaa Inclusive Education
Technical risk assessment

Predictive Maintenance Technology Risk Assessment: Sensor, Analytics and Process Controls

A worked example of how a maintenance team identifies, scores and controls the risks of condition monitoring, sensor data, analytics and workforce skills that decide whether predictive maintenance protects assets or fails them.

Updated November 2026 8 hazards assessed Review every 6 months Highest risk: Critical to Medium
See the Predictive Maintenance Technology course

Executive summary

This assessment shows how a maintenance team can find, score and control the risks that come with predictive maintenance technology. Eight risks were chosen because they are plausible for a site with connected sensors, cloud analytics and planners who act on alerts. Sensor drift, analytics errors and cyber intrusion are the most likely to affect results, while deferred maintenance on critical assets is the most serious, since a missed warning can cause injury. Each risk is scored before controls and again after existing and added controls. Most fall to Medium or Low, and none remains High or Critical. The first actions are a rule that blocks deferral of safety defects, secure remote access, sensor health checks and exit terms in vendor contracts. The site is a teaching example, so every score illustrates method, not a real plant.

Focus area
Predictive maintenance technology and asset reliability
Setting
Maintenance function of a mid-sized manufacturing site with rotating plant, connected sensors and a cloud analytics platform
Method
Five by five likelihood and impact risk register

Scope

This assessment covers condition monitoring sensors, data collection, analytics models, alert handling, work orders and the maintenance decisions they drive for a set of rotating plant assets. It covers safety, quality, continuity, data protection and cost. Jurisdiction, site size, vendor and cloud platform are assumptions made for teaching, and no real plant or supplier is described.

Context

The site runs pumps, motors, compressors and conveyors that are monitored by vibration, temperature and current sensors. Readings flow through gateways into a cloud analytics platform that scores asset health and creates alerts. Maintenance planners turn alerts into work orders, technicians carry out inspections, and the reliability engineer reviews trends each month. Each risk is scored before controls to show its natural level, then again after existing controls and the additional controls proposed here.

Objectives, assets and stakeholders

The objective is to keep critical assets available and safe, catch faults early enough to plan repairs, protect operating data and personal information, and use the maintenance budget well. Assets in scope are rotating plant, sensors and gateways, the analytics platform and its stored history, maintenance records and the vendor contracts that support them. Stakeholders include operators, technicians, planners, reliability engineers, the plant manager, IT and security staff, the finance partner, procurement, staff representatives and the vendor. Every risk has one owner role, and the plant operations manager reviews the register with the leadership team at each review.

Risk matrix: before and after controls

Each cell shows hazard numbers placed by likelihood and impact. Compare the two grids to see how controls move each hazard.

Before controls

Inherent risk

Risk before controls: rows are likelihood from almost certain to rare, columns are impact from insignificant to catastrophic. Cells list hazard numbers.
1
Insignificant		 2
Minor		 3
Moderate		 4
Major		 5
Catastrophic
5 Almost certain
4 Likely Likelihood 4, impact 2, score 8, medium: H06 Likelihood 4, impact 3, score 12, high: H01 Likelihood 4, impact 5, score 20, critical: H05
3 Possible Likelihood 3, impact 3, score 9, medium: H08 Likelihood 3, impact 4, score 12, high: H02 H04 Likelihood 3, impact 5, score 15, high: H03
2 Unlikely Likelihood 2, impact 4, score 8, medium: H07
1 Rare

After controls

Residual risk

Risk after controls: rows are likelihood from almost certain to rare, columns are impact from insignificant to catastrophic. Cells list hazard numbers.
1
Insignificant		 2
Minor		 3
Moderate		 4
Major		 5
Catastrophic
5 Almost certain
4 Likely
3 Possible
2 Unlikely Likelihood 2, impact 2, score 4, low: H06 H08 Likelihood 2, impact 3, score 6, medium: H01 H02 H04 Likelihood 2, impact 4, score 8, medium: H03 H05
1 Rare Likelihood 1, impact 3, score 3, low: H07
  • Low 1 to 4
  • Medium 5 to 9
  • High 10 to 16
  • Critical 17 to 25

Hazard register

A summary first, then the detail for each hazard.

Summary of hazards with scores before and after controls
No.HazardBeforeAfterOwner
H01 Sensor data quality drift and calibration errors 12 High 6 Medium Reliability engineer
H02 False alarms and missed failures from analytics errors 12 High 6 Medium Data analytics lead
H03 Cyber intrusion into connected sensors and telemetry 15 High 8 Medium IT security manager
H04 Dependence on a single sensor vendor and cloud platform 12 High 6 Medium Procurement manager
H05 Deferred maintenance on critical assets due to over-reliance on predictions 20 Critical 8 Medium Plant operations manager
H06 Skills gap in interpreting condition data 8 Medium 4 Low Training coordinator
H07 Worker privacy and misuse of monitoring data 8 Medium 3 Low Privacy and legal adviser
H08 Investment shortfall and weak return on monitoring spend 9 Medium 4 Low Finance business partner
H01

Sensor data quality drift and calibration errors

Technology and data Before 12 After 6

Sensors drift, are mounted loosely or lose calibration, so readings look normal while a machine degrades, or false trends trigger unnecessary work on healthy equipment.

Who or what is affected: Maintenance planners, reliability engineers and production staff

Causes

Vibration loosening mounts, temperature drift, damaged cables and calibration left to a yearly visit.

Consequences

Missed failures, unplanned breakdowns and wasted work on healthy equipment.

Existing controls

  • Sensors are calibrated once a year by the vendor.
  • Planners check readings against a manual route inspection each month.

Control gaps

Sensor health is not monitored separately from asset health, so faulty data looks like real data.

Additional controls

  • Prevent: Fit mounting checks and tag each sensor with a calibration due date in the asset system.
  • Detect: Flag flat-line readings and sudden step changes automatically for review within one shift.
Likelihood 4 x impact 3 before; 2 x 3 after Owner: Reliability engineer Target: Within 60 days Early warning: A sensor reporting no change for 24 hours or a calibration past its due date.
H02

False alarms and missed failures from analytics errors

Operational Before 12 After 6

Alert thresholds or models trained on narrow data raise too many false alarms or miss developing faults, so teams either ignore alerts or act on noise and miss real warnings.

Who or what is affected: Planners, technicians and plant managers

Causes

Generic thresholds, few recorded failures and models not retrained after equipment changes.

Consequences

Alert fatigue, missed faults, unnecessary parts replacement and loss of trust in the system.

Existing controls

  • Alert thresholds are set by the vendor from standard settings.
  • A planner reviews each alert before a work order is raised.

Control gaps

Model performance is reviewed only after a failure, not on a routine schedule.

Additional controls

  • Prevent: Tune thresholds for each asset class using the site's own failure history and record each change.
  • Detect: Measure alert precision monthly by comparing alerts with confirmed faults found at inspection.
Likelihood 3 x impact 4 before; 2 x 3 after Owner: Data analytics lead Target: Within 90 days Early warning: Alert precision below the agreed level for two consecutive months.
H03

Cyber intrusion into connected sensors and telemetry

Technology and data Before 15 After 8

Gateways, sensors or the cloud account are compromised through weak passwords, unpatched devices or phishing, so monitoring data is altered or stolen, or monitoring stops entirely.

Who or what is affected: Operations, maintenance staff, the IT team and suppliers

Causes

Default credentials, unpatched firmware and remote access shared between vendors.

Consequences

Falsified health data, production stoppage, data loss and a regulatory notice if personal data is exposed.

Existing controls

  • Gateways sit on a separate network segment from office systems.
  • Cloud accounts require a password and are reviewed by IT each year.

Control gaps

Vendor remote access is not logged or time limited, and firmware updates are not tracked.

Additional controls

  • Prevent: Enable multi-factor sign in for all platform accounts and change default device credentials before commissioning.
  • Respond: Isolate affected gateways and switch critical assets to scheduled inspection until the network is confirmed clean.
Likelihood 3 x impact 5 before; 2 x 4 after Owner: IT security manager Target: Within 30 days Early warning: Repeated failed sign in attempts or a gateway reporting unknown firmware.
H04

Dependence on a single sensor vendor and cloud platform

Supplier Before 12 After 6

The site relies on one vendor for sensors, software and hosting, so a price rise, service failure or product withdrawal leaves monitoring without support or access to its own data.

Who or what is affected: Procurement, maintenance and finance teams

Causes

Lock in from a bundled purchase, proprietary data formats and limited internal skills.

Consequences

Loss of monitoring, higher costs, costly migration and delayed repairs.

Existing controls

  • The contract includes a service level for response to faults.
  • Historical data is exported to the site server each week.

Control gaps

Exit terms are not tested, and the weekly export has never been restored to check it works.

Additional controls

  • Prevent: Write data ownership, export formats and exit support into all new supplier contracts.
  • Detect: Review vendor performance and financial health each quarter using a short scorecard.
Likelihood 3 x impact 4 before; 2 x 3 after Owner: Procurement manager Target: Within 90 days Early warning: A vendor service response time breached twice within one quarter.
H05

Deferred maintenance on critical assets due to over-reliance on predictions

Safety Before 20 After 8

A planner defers a repair because the model shows low risk, while a fault such as bearing failure, overheating or a guard defect develops and leads to injury or fire.

Who or what is affected: Maintenance technicians, machine operators and nearby workers

Causes

Pressure to keep production running, trust in predictions and unclear deferral rules.

Consequences

Serious injury, fire, catastrophic equipment failure and enforcement action.

Existing controls

  • Critical asset alarms override planner deferrals.
  • Safety inspections follow a fixed schedule for guards and pressure parts.

Control gaps

Deferral rules do not cover safety-related defects, and risk acceptance is not recorded.

Additional controls

  • Prevent: Make safety-related defects non-deferrable in the work order system unless a competent person signs a recorded risk acceptance.
  • Respond: Stop or isolate equipment when a critical indicator and a visible defect coincide, and call the supervisor at once.
Likelihood 4 x impact 5 before; 2 x 4 after Owner: Plant operations manager Target: Within 14 days Early warning: Deferred work orders on safety-related assets older than two weeks.
H06

Skills gap in interpreting condition data

People Before 8 After 4

Technicians and planners lack the training to read vibration spectra, trends and alerts, so they misread data, act late or ignore warnings that matter.

Who or what is affected: Technicians, planners and supervisors

Causes

High staff turnover, rapid rollout of tools and training given only once.

Consequences

Wrong decisions, extra repair cost and poor use of the system.

Existing controls

  • New technicians attend a vendor introduction course.
  • A reliability engineer reviews major alerts with the team each week.

Control gaps

Competence is not checked after the first course, and no one tracks skill decay.

Additional controls

  • Prevent: Set a competence matrix for each role and require sign off before a person acts on analytics alone.
  • Recover: Run a short refresher after any missed fault and record the lesson learned.
Likelihood 4 x impact 2 before; 2 x 2 after Owner: Training coordinator Target: Within 60 days Early warning: Alerts closed without a recorded action or repeated wrong fault diagnoses.
H07

Worker privacy and misuse of monitoring data

Legal and compliance Before 8 After 3

Sensors or software record operator location, pace or personal identifiers, and the data is used to judge individuals beyond the purpose agreed with staff, creating legal and trust problems.

Who or what is affected: Operators, technicians and their representatives

Causes

Vague purpose, wearables or cameras added without review and reuse of data for performance reviews.

Consequences

Complaints, disciplinary disputes, legal action and reputational harm.

Existing controls

  • Asset data is linked to machines rather than named people by default.
  • Access to the platform is limited to maintenance and reliability staff.

Control gaps

No review happens when new data fields are added to the platform.

Additional controls

  • Prevent: Write a purpose statement for monitoring data and agree it with staff representatives before use.
  • Detect: Audit platform user lists and exports each quarter for any personal data fields.
Likelihood 2 x impact 4 before; 1 x 3 after Owner: Privacy and legal adviser Target: Within 60 days Early warning: Monitoring data requested for individual performance reviews.
H08

Investment shortfall and weak return on monitoring spend

Financial Before 9 After 4

Sensors, licenses and integration cost more than planned, while avoided failures are hard to prove, so the program loses funding or is cut before it delivers measurable value.

Who or what is affected: Finance business partner, site leadership and maintenance teams

Causes

Optimistic benefit estimates, hidden license fees and scope creep.

Consequences

Stalled program, wasted spend and loss of management support.

Existing controls

  • Business cases list the expected avoided downtime costs.
  • The annual budget is reviewed by the site leadership team.

Control gaps

Benefits are not tracked against the original case after go live.

Additional controls

  • Prevent: Phase rollout by asset criticality and release each phase only when the previous one meets its agreed measures.
  • Respond: Re-plan the scope and drop low value sensors if benefits fall below the approved case.
Likelihood 3 x impact 3 before; 2 x 2 after Owner: Finance business partner Target: Within 90 days Early warning: Program cost forecast above the approved budget tolerance.

Rating method and evaluation

Each risk is rated on two five-point scales. Likelihood runs from Rare, meaning unlikely over the life of the system, to Almost certain, meaning expected to occur. Impact runs from Insignificant to Catastrophic, covering safety harm, lost production, data loss, cost and trust. Multiplying the two gives a score from 1 to 25 in one of four bands: Low from 1 to 4, Medium from 5 to 9, High from 10 to 16 and Critical from 17 to 25. Inherent risk is the score before any controls, judged as if existing controls were absent. Residual risk is the score after existing controls and the additional controls proposed here, and it must never be higher than inherent. Safety impacts are never rated down only because no harm has yet occurred. Critical risks need immediate action and senior sign off, High risks need a dated plan, Medium risks need routine controls with a named owner, and Low risks are accepted and watched. Ratings are challenged by the people who run the assets in a short review at each cycle, and any rating is revisited when a monitoring indicator reaches its escalation threshold.

Likelihood, impact and rating bands

Likelihood

  • 5 Almost certain: Expected to happen in most circumstances.
  • 4 Likely: Will probably happen in many circumstances.
  • 3 Possible: Could happen at some time.
  • 2 Unlikely: Not expected, but it could happen.
  • 1 Rare: Only in exceptional circumstances.

Impact

  • 1 Insignificant: No lasting effect. Handled within routine work.
  • 2 Minor: Small effect that is quickly corrected.
  • 3 Moderate: Noticeable effect that needs managed effort to correct.
  • 4 Major: Serious effect on people, delivery or reputation.
  • 5 Catastrophic: Severe or lasting effect that threatens the whole activity.

Bands (likelihood x impact)

  • Low (1 to 4): Accept and monitor.
  • Medium (5 to 9): Manage with routine controls.
  • High (10 to 16): Senior attention and a treatment plan.
  • Critical (17 to 25): Act first. Do not proceed until reduced.

Monitoring, escalation and review

Review this assessment every 6 months, and sooner when an indicator below reaches its threshold.

Early-warning indicators and escalation thresholds
Indicator to watchEscalate when
Sensors reporting flat or stepped readings, or past their calibration due date.Escalate to the reliability engineer within one shift, and to the maintenance manager if unresolved after a week.
Alert precision for each asset class, measured monthly against confirmed faults.Escalate to the data analytics lead when precision stays below the agreed level for two months, and to the plant manager after a third.
Failed sign in attempts or unknown firmware on gateways.Escalate to the IT security manager at once, and isolate the device until it is verified.
Safety-related work orders deferred beyond their agreed date.Escalate to the plant operations manager at once, and stop the asset if a visible defect is present.
Vendor service breaches or price changes recorded in the quarter.Escalate to the procurement manager within five working days, and to the finance partner if costs exceed the approved budget.

Contingency arrangements

If a sensor network is compromised or a deferred safety defect is found, the plant operations manager stops the affected asset and isolates the network segment. The IT security manager preserves logs, removes vendor access and restores gateway settings from a tested copy. Maintenance moves critical assets to scheduled inspection until the data is trusted again. Legal and privacy advisers are contacted for any personal data exposure, and specialist review is needed before any regulator is informed. Staff are told early and plainly what is changing.

Prioritized action plan

  1. Make safety-related defects non-deferrable without a recorded sign off by a competent person, owner plant operations manager, within 14 days.
  2. Secure remote vendor access and change default device credentials, owner IT security manager, within 30 days.
  3. Check each sensor against its calibration due date and health record, owner reliability engineer, within 60 days.
  4. Write data export and exit support terms into vendor contracts, owner procurement manager, within 90 days.
  5. Set a competence matrix and refresher cycle for every analytics user, owner training coordinator, within 60 days.

Conclusion

Predictive maintenance pays off only when data is trustworthy, alerts lead to sound decisions and people can act on them. The highest scores sat with deferred safety work, cyber intrusion and sensor drift, and each has controls that lower the residual score. Treat the register as a living record, review it every six months or after any major change, and seek qualified engineering, safety, privacy and legal advice for any real site decision.

Assumptions and limitations

Assumptions

  • The site has a computerized maintenance management system that holds asset records and work orders.
  • Sensors are installed on critical rotating assets and are maintained by the site team and an external vendor.
  • Jurisdiction is not specified, so legal and privacy duties are described in general terms only.
  • Analytics models were built with historical failure records, which are assumed to be complete.

Limitations

  • The scores are examples for learning and do not describe any real plant, vendor or platform.
  • Safety-critical equipment and personal data rules need specialist review, and this content does not replace a formal site-specific assessment or licensed engineering advice.
  • Sensor and analytics performance varies by equipment type, so no accuracy figures are given.
  • This is educational content only and is not legal, privacy or engineering advice.

Related frameworks to study

  • Failure mode and effects analysis
  • Hierarchy of controls
  • Plan, do, check, act improvement cycle
  • Reliability centered maintenance review

Technical Risk Assessment

China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.

For Learning Reference Only: This assessment demonstrates how risk assessment principles are applied in a practical setting, providing a useful reference for learning and understanding. Actual assessments may be adapted to suit specific site conditions and circumstances.

Keep learning

Study the subject behind this assessment.

Predictive Maintenance Technology course Read the best practices for this course

More Technical risk assessments

Hourly Active: 45,552 Learners