MechSignal is designed to bring equipment readings, flags and inspection outcomes into one place, so maintenance teams can see what’s changing and decide what to check first.
Each capability is in development and will be refined with pilot partners. Nothing here describes a released, validated product.
A list of monitored assets grouped by site, line or area, each with a plain-language status: within baseline, watch, review suggested, or baseline building when there isn’t yet enough data to judge.
Trend charts for the signals your equipment provides — such as vibration velocity, bearing or housing temperature, pressure and motor current — drawn against the machine’s own learned baseline range.
Models that identify readings deviating from learned operating behaviour. Each flag links to the readings that triggered it so a person can judge whether it matters.
A flag means “this looks different from usual” — it is not a diagnosis, and it does not predict when or whether a machine will fail. Detection quality depends on the data available and must be validated for each machine type.
A ranked list of suggested inspections that weighs how far readings have moved from baseline, how long the deviation has lasted, the criticality your team assigns to each asset, and recent maintenance history.
After an inspection, technicians record the outcome — issue confirmed, no issue found, or needs more data — with notes and photos. That record stays with the asset.
Feedback is used to review how flags are raised for each machine, so the system can be tuned to what your team actually finds in the field.
A per-machine timeline of readings, flags, inspections and outcomes, plus summary reports for maintenance reviews and shift handovers.
Connect existing historians, PLCs or gateways, or add sensors where none exist. Import maintenance records where available.
Use a period of operating history to learn typical behaviour per machine and operating mode. Assets show “baseline building” until this is ready.
Compare new readings to the baseline. Sustained or significant deviations are flagged and ranked for review.
Flags become inspection tasks. Technicians record findings, which feed back into how flags are tuned.
This is the planned design, not a deployed system. Services and data flow may change as development continues and as pilot requirements become clear.
Existing sensors, PLCs, historians or new sensors, via an edge gateway.
Collects equipment data and organises it into asset models (site, line, machine, signal).
Holds longer-term readings and maintenance records used for baselines, training and reports.
Trains, evaluates and runs per-asset anomaly models; results are written back for review.
Sends flag notifications by email, SMS or integrations to subscribed maintenance contacts.
The design reads data only. MechSignal is not planned to write to PLCs or control equipment.
Deployment options, data residency and access controls will be agreed with each pilot partner. [Data handling policy — to be published]
Network access, gateway configuration and permissions will be reviewed with your IT/OT teams before any connection is made.
Monitoring is only as good as the data behind it. We’ll assess these together before agreeing to a pilot.
The machines in scope need relevant measurements available — for example vibration, temperature, pressure, flow or motor current — through one of:
If a machine has no relevant signals, MechSignal can’t monitor it.
Baselines need representative data covering the machine’s usual loads and operating modes. How much is needed depends on the equipment and how variable its duty cycle is. Until a baseline is ready, the asset is shown as “baseline building”.
Model performance has to be validated for the specific machines and operating conditions in scope. During a pilot, flags are compared against technician findings to measure how useful they are — including missed issues and false alarms — before anyone relies on them.
We don’t publish generic accuracy or downtime figures, because they wouldn’t reflect your equipment.
Technicians and planners need time to review flags and record outcomes. That feedback is how the system is checked and improved, so a named maintenance contact is part of every pilot.
Tell us about your machines and data sources. We’ll give you a straight answer on whether a pilot makes sense.