Service 02
Translation Workflow Integration
Assisted translation for recurring business documents, with a maintained terminology base and a defined human review stage for anything customer-facing or contractual. Seven weeks.
¥33,000
View Service
Service 01 — ¥45,000 JPY — 12 Weeks
Know what your equipment
is telling you
Condition monitoring built from the sensor output your equipment is already producing. Fault signals reach the right person before a scheduled inspection would have caught anything.
What This Delivers
Earlier signals. Fewer surprises.
Faults caught earlier
Developing conditions in your equipment become visible days or weeks before they would register on a scheduled inspection cycle. Your maintenance team decides what to do with that time.
Alert routing that fits your team
Alerts reach the people who need to act on them, through the channels they already check. No new interfaces for staff to learn.
Thresholds owned by internal staff
By the end of the twelve weeks, your maintenance team understands the thresholds and can adjust them as conditions change. No ongoing dependency on us.
The Situation
Equipment is already signalling. Most of it goes unread.
Modern production equipment — even equipment that has been running for years — produces a steady stream of operational data. Temperature, vibration, pressure, cycle time, current draw. Most of it is logged and then never looked at until something stops.
Scheduled maintenance visits cover what they cover. They can't account for what changed between visits. When a fault develops in the interval, the first indication is often an unexpected stop or a part that has worn further than it should have.
The data to have caught it earlier was there. It just wasn't being read in a way that connected to the maintenance team's workflow.
Common Patterns We See
-
—
Sensor logs exist but are reviewed only after a stoppage, when the pattern is too late to act on.
-
—
Maintenance intervals are fixed by calendar rather than by what the equipment is doing between visits.
-
—
Alert systems exist but generate enough false signals that staff have learned to treat them as background noise.
-
—
Maintenance history is held in paper logs or disconnected spreadsheets, so pattern recognition depends on individual memory.
Our Approach
Condition monitoring from data you already have
We don't start by proposing new sensors or platforms. We start by looking at what your equipment is already outputting and what your maintenance team already knows about how it behaves.
Data Collection
Work from existing sensor output
We review what your equipment is logging and identify which signals are most useful for detecting the fault types your maintenance history shows. No hardware additions required at this stage.
Threshold Definition
Set thresholds with your maintenance team
Thresholds are defined collaboratively. Your team's knowledge of what the equipment normally does in different conditions shapes what gets flagged and what doesn't. We document the reasoning behind every threshold.
Alert Routing
Connect signals to the right workflow
We route alerts into whatever communication or task system your maintenance team is already using. The goal is that a signal reaches the right person without requiring them to check a new tool.
Review Process
Track correct and incorrect alerts together
We build a review process for both accurate alerts and false ones. False alert rates are reported alongside detection rates — both numbers matter, and both are visible from the start.
What the Engagement Looks Like
Twelve weeks, structured by observation
Weeks 1–3
Data review and scope
We review sensor output and maintenance history with your team. Together we identify which equipment and which fault types are the priority. Deliverables and assumptions are agreed in writing.
Weeks 4–9
Build, calibrate, and observe
The monitoring system is built and connected to your alert channels. We run it in parallel with your existing inspection schedule so nothing changes for your team while we observe real conditions. Thresholds are adjusted as we learn more.
Weeks 10–12
Handover and close
Threshold adjustment procedures are handed to your maintenance team. We document the logic behind every decision made during calibration. Your team closes the engagement able to operate without us.
Investment
What this engagement covers
One fixed price for the full twelve weeks. No separate charges for calibration rounds or documentation. The scope is agreed before work begins.
What's Included
-
—
Data review and sensor output assessment
-
—
Threshold definition with your maintenance team
-
—
Alert routing into your existing workflow
-
—
Review process for correct and false alerts
-
—
False alert rates reported alongside detection rates
-
—
Existing maintenance schedules retained throughout
-
—
Threshold adjustment procedures handed to internal staff
-
—
Full documentation of calibration decisions and assumptions
Load Table — Indicative Figures
| Metric | Before | After Implementation |
|---|---|---|
| Fault detection lead time | At or after inspection interval | Days to weeks before threshold breach |
| Maintenance team review time (per equipment unit) | ~2 hours per scheduled inspection | ~20 min per alert reviewed |
| Alert volume (early calibration) | — | Higher; reduces after threshold tuning in weeks 6–9 |
| False alert rate (post calibration) | Not tracked | Tracked and reported each week |
Assumptions: Equipment already producing sensor output at 1-minute or shorter intervals. Maintenance history available covering at least 12 months. Alert routing assumes an existing messaging or task system is in use. Lead time improvement depends on fault type and sensor coverage; not all fault types will produce early signals from existing sensors alone.
How We Measure Progress
What gets tracked and how results are defined
01
Detection rate
Proportion of fault events that produced an alert before the fault was identified by other means. Reported weekly.
02
False alert rate
Proportion of alerts that did not correspond to an actual developing fault. Both numbers appear in every report, side by side.
03
Lead time
How far ahead of a fault the alert arrived. Measured per event and tracked across the observation window.
04
Threshold stability
How often thresholds required adjustment after initial calibration. An indicator of how well the initial setup accounted for real conditions.
Note on expectations: Twelve weeks includes a calibration period during which false alert rates will be higher than at close. This is expected. We do not represent the figures at close as a guaranteed ongoing rate — equipment behaviour changes, and threshold adjustment is a continuing responsibility handed to your team.
Our Commitment
Scope delivered or we talk about it
The scope of this engagement — data review, threshold definition, alert routing, calibration observation, and handover — is agreed in writing before we begin. If any part of that scope is not delivered, that's a conversation we're prepared to have.
We don't offer result guarantees against operational metrics we don't control. Equipment behaviour, sensor coverage, and fault frequency all affect what the system can detect. What we can stand behind is our process and our documentation.
If you'd like to understand what's possible before committing, we're available for a preliminary conversation with no obligation.
Maintenance Panel — After Close
Threshold adjustment
Handled by your maintenance team using the documented procedure. Expected as equipment ages or operating conditions change.
Responsible: Internal maintenance staff
Alert review log
Maintained by whoever receives alerts. Correct and false alerts are logged to allow periodic threshold review.
Responsible: Alert receiver, reviewed monthly
Sensor connection
If sensor output format changes due to equipment replacement or firmware update, the connection may need reconfiguration.
Responsible: IT or maintenance staff; we document the connection for this purpose
Adding equipment
The process for adding new equipment to the monitoring scope is documented. This can be done by internal staff or we can assist under a separate scope.
Responsible: Internal staff; separate engagement available
Getting Started
The path from here
Step 01
Send us a brief note
Describe your equipment situation — what's running, what data it produces, and what your current maintenance arrangement looks like. No detailed brief required at this point.
Step 02
We respond within two business days
We'll let you know whether this service fits your situation and, if so, what a data review would involve. If it's not a fit, we'll say so.
Step 03
Scope agreed, work begins
If we proceed, scope and assumptions are set in writing before any data is reviewed. The twelve-week engagement follows the structure outlined above.
Service 01 — Predictive Maintenance Setup
¥45,000 — Twelve Weeks
If your equipment is already producing operational data and your maintenance team is working from scheduled inspection intervals, it may be worth a conversation about what condition monitoring could add.
Send a MessageOther Services
Explore other areas where we work
Service 03
Operations Dashboard Build
A live operational view drawing production, inventory, and order figures into one place, replacing figures currently assembled by hand each morning. Eight weeks.
¥37,000
View Service