Craftsman Information Device Operation cheat sheet

A craftsman information device stores three layers: input capture (forms, scans, sensor reads), local records, and synced master data; loss at any layer breaks traceability.

Binary units: 1 byte = 8 bits; 1 KB = 1,024 bytes; 1 MB = 1,024 KB; 1 GB = 1,024 MB. Storage specs usually quote decimal (1 GB = 1,000,000,000 bytes), so usable capacity reads lower.

Field data types: text, integer, decimal, date/time, boolean, and enumerated list. A value stored in the wrong type is a data-integrity fault, not a display fault.

Record identity rule: every record needs a unique key (serial, job number, or generated ID). Duplicate keys cause overwrites during sync and are unrecoverable without an audit log.

Timestamps should be stored in UTC with a local offset shown at display; mixing local-time records across daylight-saving changes creates ordering errors.

Read/write media classes: internal flash (fast, finite write cycles), removable SD/USB (portable, corruption risk on removal), and network share (shared, latency-bound).

Offline-first design: queue writes locally, mark each with a sync state (pending, sent, confirmed), and reconcile on reconnect; unmarked queues cause silent duplicates.

Data classification tiers (public, internal, confidential, restricted) determine encryption and sharing rules; a device holding restricted data must encrypt at rest.

Audit trail minimum: who changed what, old value, new value, timestamp, and device ID. Without the old value, corrections cannot be reversed.

Pre-deployment checklist order: charge to 100%, install SIM/memory card, power on, set locale and time zone, join network, enroll in management, then verify sync with a test record.

Factory reset erases local records and credentials but not server-side data; export or confirm sync before resetting any device holding unsynced work.

Time zone and clock source: enable automatic network time; a manual clock drift over 2 minutes breaks timestamp ordering and token-based sign-in.

Firmware update rule: apply updates only on a charged battery above 50% or on mains power, and never interrupt the write phase; an interrupted flash can brick the unit.

Network configuration: DHCP for routine use, static address only where the site requires it; record the assigned address, subnet, gateway, and DNS on the device label.

Enrollment/management profiles push Wi-Fi, certificates, and app settings. A profile that fails to apply usually indicates an expired certificate or wrong enrollment token.

Permission model: assign the least role that allows the task (operator, supervisor, administrator); administrator accounts are for configuration only, not daily data entry.

Calibration and pairing: sensors and peripherals must be paired and zeroed at setup, and the calibration date recorded, before the device is released to production.

Acceptance test before handover: capture one record, sync it, retrieve it on the server, and confirm the audit entry shows correct user, device, and timestamp.

Start-of-shift sequence: confirm charge or swap battery, verify network and sync status, check pending queue is empty, then open the assigned work list.

End-of-shift sequence: submit open records, force a sync, confirm zero pending items, log out, and place the device on charge in its assigned location.

Work-order states typically run assigned, in progress, paused, complete, closed; skipping a state or closing with open sub-tasks creates reconciliation gaps.

Battery rule of thumb: lithium cells degrade faster when kept below 20% or above 80%; mid-shift top-ups are safer than full discharge cycles.

Barcode/QR scanning: hold 10-25 cm from the label, keep the beam perpendicular, and clean the window with a dry lint-free cloth; repeated no-reads usually mean a damaged or low-contrast label.

Handover between operators requires a written or in-app note of open items, device condition, and pending syncs; verbal-only handover loses exceptions.

Exception handling: when a step cannot be completed, record the reason code and photo evidence at the point of failure rather than reconstructing it later.

Device sharing: log out before passing the device on; records inherit the logged-in user, so shared logins destroy accountability in the audit trail.

Shift reports should reconcile count of records captured against count confirmed on the server; a mismatch indicates a queue or connectivity fault.

Double-entry verification: for high-consequence fields (serial, quantity, measurement), enter twice or scan instead of typing; a mismatch blocks submission.

Range and format validation catch most entry errors: set min/max, required fields, and pattern masks (for example, 8-digit asset numbers) at the form level.

Decimal and unit discipline: record the unit with every measurement (mm, kg, °C) and never mix unit systems in one field; unit errors are a common recall cause.

Search strategy: start with the unique key (serial or job number), then narrow by date range, then by free text; free-text-only searches return false matches.

Transcription rule: read back numeric strings digit by digit and confirm check digits; spoken 'one-five' versus 'fifteen' ambiguity causes field errors.

Correction protocol: never overwrite a submitted value; issue a correction record referencing the original, with reason and approver, so history stays intact.

Legibility standards: print in block capitals, use a black or blue pen, and strike through errors with a single line plus initials; correction fluid is prohibited on records.

Retrieval verification: confirm the record you opened matches the intended key and date before acting on it; look-alike serials differing by one character are a known trap.

Backlog rule: clear the pending queue before the device sleeps or the shift ends; queued records older than the retention window may be rejected by the server.

Fault-isolation order: power and charge, then physical connections and card seating, then network, then application, then account permissions; fix the lowest layer first.

No power: charge 30 minutes on a known-good cable and charger, then try a forced restart (hold power 10-15 seconds); no response after both indicates battery or hardware failure.

Will not sync: check Wi-Fi/cellular signal, confirm the server address and certificate validity, verify the account is not locked, then retry; record the error code before clearing it.

Scanner no-read: clean the window, check label contrast and damage, test with a known-good label, then recalibrate; if a known-good label also fails, escalate the scanner.

Slow performance: close background apps, free storage below 10% remaining, restart, and clear the app cache; persistent slowness after these points to failing storage.

Touch or display faults: remove screen protector bubbles, clean with a dry cloth, restart, and test in a diagnostic screen; unresponsive zones after restart are hardware faults.

Preventive maintenance schedule: daily clean and charge check, weekly cable and connector inspection, monthly backup verification and storage cleanup, quarterly calibration check.

Escalation package: device serial, firmware version, error code, time of fault, steps already tried, and whether data is at risk; missing details delay repair.

Never attempt board-level repair or open a sealed battery compartment in the field; that voids warranty and creates a safety hazard.

Lithium battery handling: no puncturing, crushing, or charging below 0 °C; a swollen, hot, or leaking battery is removed to a fire-resistant area and reported, not used.

Electrical safety: inspect charger cables for damage before use, keep chargers off wet floors, and never charge in a classified hazardous atmosphere unless the device is rated for it.

Intrinsically safe (Ex-rated) devices must not be substituted with standard units in hazardous zones, and their certification label must remain legible and intact.

Record retention: regulated records are kept for the period set by the applicable standard or contract; deletion before that date is a compliance breach even if the data is obsolete.

Personal data handling: collect only what the task requires, restrict access by role, encrypt in transit and at rest, and report suspected loss or theft within the site's incident window.

Incident reporting: any data loss, device loss, injury, or near miss is logged with time, location, people involved, and immediate action taken, before the shift ends.

Sign-off authority: only personnel with current authorization may approve or close regulated records; expired training invalidates the signature.

Workstation ergonomics and site rules: follow site PPE, permit-to-work, and lone-working rules; device use never overrides a site safety instruction.

Disposal and decommissioning: wipe data to the required standard, remove memory cards and SIMs, and route batteries and electronics to approved recycling streams.