How do you translate ISO/TS 15143-3:2020 earth-moving machinery telematics reports so that a construction manager reads the same machine identifier, GPS position, timestamp, fuel figure and operating-hour value in every language? A fleet dashboard may look like a series of short labels, yet the language beneath each value controls maintenance decisions, worksite planning, ownership checks and customer reporting. ‘Operating hours’, ‘engine hours’, ‘idle time’, ‘fuel used over 24 hours’ and ‘lifetime cumulative fuel’ are not automatic synonyms.
This world-facing guide teaches translators, heavy-equipment operators, fleet-data integrators, contractors, OEM documentation teams and learners how to localise machine telematics summaries, API field descriptions, equipment-status exports, GPS histories and mixed-fleet reports without rewriting the underlying measurements. The examples are entirely fictional and designed to expose real classes of translation error: units silently converted, locations presented as live, asset identifiers translated as words, missing values mistaken for zero and cumulative counters confused with short-period consumption.
Quick answer: Preserve machine and equipment identifiers exactly; keep each reading attached to the original measurement definition, unit, source system and timestamp; distinguish latest known from current location; never derive engine operation from an unrelated status; and validate a translated dashboard against the untouched machine data before release. When OEM terminology differs, resolve it in a controlled glossary instead of forcing every vendor field into a misleading generic label.
- Primary search intent: translating earth-moving equipment telematics data and reports.
- Protect machine IDs, PIN/VIN strings, serial numbers and field keys.
- Keep accumulated operating hours separate from elapsed hours in a reporting window.
- Make last-known GPS positions and update age clear; do not invent real-time tracking.
- Preserve quantity units, missing-value states, source provenance and firmware-specific definitions.
What ISO/TS 15143-3:2020 actually addresses
The official ISO catalogue describes ISO/TS 15143-3:2020 — Earth-moving machinery and mobile road construction machinery — Worksite data exchange — Part 3: Telematics data as a technical specification for providing mobile-machine status data from a telematics provider’s server to a customer application through the internet. It addresses the communications requests and responses for specified data elements. It is not the on-board sensor collection protocol, a CAN-bus standard or the wireless link between the machine and provider server. The 2020 edition was confirmed in 2023, and a successor draft is in development as of November 2026.
An important boundary follows immediately: a translator may work with reports described as ‘AEMP’ or ‘ISO-compatible’, but the actual delivered fields depend on what the provider, customer API and machine support. The ISO catalogue does not guarantee that every vehicle transmits every possible field, that a displayed reading is accurate, or that the connection is live. Refer to the official ISO/TS 15143-3:2020 catalogue entry and the controlling technical specification when checking compliance. The examples below do not reproduce a licensed schema or certify interoperability.
First understand the data path
Picture an excavator, a telematics unit, a provider server, an API service and a contractor’s multilingual dashboard. The machine and provider may record a value at one moment, store it, then deliver it later when the client requests a status. The language team normally sees the final user interface or an exported dataset, not the measurement pipeline. A displayed latitude and longitude could describe where the machine was at the time of the last received position. It may not show where the machine is now. A fuel counter could be cumulative since manufacture or reset. The source definition matters more than how familiar ‘fuel used’ sounds.
A second question concerns software labels versus machine data. A machine ID, structured field key, model code or service endpoint must generally remain stable across languages so systems can match records. The visible label beside it can be translated, provided the meaning is preserved. That separation—protect the data, localise the explanation—is fundamental to trustworthy engineering translation and to the narrow search intent this page owns.
Terminology that changes fleet decisions
Equipment ID
What the field means. The identifier under which a machine is represented in a customer’s or provider’s data collection. It may be an internal asset tag rather than a manufacturer serial number. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not replace the asset tag with a translated model description or assume two systems use the same identifier namespace. Confirm its field name, scope, format, leading zeros and relation to any serial, PIN or VIN. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Manufacturer serial number
What the field means. A producer-assigned machine identity, which may be used by service, warranty or supply records. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Changing character case, omitting a prefix or inserting spaces can break exact-match lookup in an external system. Protect original string and compare it with registration or machine documentation; never transliterate as an ordinary noun. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
PIN or VIN
What the field means. A product identification number or vehicle identification number, depending on equipment and regulatory context. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. The acronym’s presence does not prove that all machinery uses a road-vehicle VIN, nor that its format matches a particular country. Retain exact original code and clarify which identity field the source API uses. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Machine make and model
What the field means. The manufacturer brand and model designation, not a free-form description of machine capability. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Translating model codes or upgrading a model name with a feature it does not guarantee can change procurement assumptions. Separate brand from localisable category labels and confirm manufacturer nomenclature. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Telematics provider
What the field means. The service supplying remote machine data, often distinct from the machine manufacturer and the customer’s fleet software. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not label provider-server values as direct physical sensor readings unless the provenance confirms that route. Record provider name, endpoint or product version and collection limitations. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
API endpoint
What the field means. The machine-readable resource a client calls to request a particular data payload or history. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not translate literal URL paths, keys, parameter names or enumerated values within an integration contract. Localise user-facing descriptions only and test the untouched endpoint using its approved documentation. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Last known location
What the field means. The most recent position available in the provider record at the time it was retrieved. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. This phrase does not mean ‘current position’ or imply that the machine has remained there since the last sample. Show source timestamp, latitude, longitude and update age where the application supports them. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Location timestamp
What the field means. The time associated with a reported position, which can differ from the time a server delivered it. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Mixing event time with API response time may produce a false sequence of machine movements. Preserve timezone offset or UTC convention, precision and field definition. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Latitude and longitude
What the field means. Numeric angular coordinates representing a location under the relevant mapping conventions. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Translating decimal separators, swapping coordinate order or stripping a negative sign can move a point far from the worksite. Test map placement, coordinate order, unit, sign, datum assumptions and format accepted by the target application. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Altitude
What the field means. A reported vertical location coordinate with a specified unit or reference, where available. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Calling it height above ground without a defined reference changes the spatial meaning. Preserve source unit and any supplied vertical datum; do not guess absent metadata. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Cumulative operating hours
What the field means. A lifetime or accumulated operating-time counter measured under the source system’s definition. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not call a 4,350-hour counter ‘hours used this week’; it may instead represent years of operation. Retain measurement date, unit, potential resets and distinction between engine and operating definitions. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Engine hours
What the field means. A measure tied to engine running under the supplier’s definition, not necessarily productive machine activity. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not substitute job hours, billed hours or hours under load. Idle time may be included depending on source conventions. Check original field definition and whether the dashboard changes labels by machine type. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Idle hours
What the field means. An accumulated or period-specific measure related to machine idling as defined by the provider or OEM. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. An engine-on state is not by itself a universal, comparable idle-hours measure across manufacturers. Check time base, sampling method, period and whether idle time is available or merely inferred. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Cumulative fuel used
What the field means. An aggregate fuel-consumption amount over a longer accumulation boundary, rather than fuel level in the tank. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Confusing litres consumed with litres remaining reverses the operational implication and may distort efficiency analysis. Preserve counter scope, unit, reset conditions and measurement timestamp. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Preceding-24-hour fuel used
What the field means. Consumption associated with a specified rolling or preceding interval, when provided in that form. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Do not confuse it with cumulative fuel or automatically convert to a calendar-day total. Verify the 24-hour window’s ending time and whether the field is an alternative to another consumption measure. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Fuel level
What the field means. Remaining fuel measure, potentially a percentage or volume, depending on the provider’s published field. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. A percentage is not a litre value and should not become a guaranteed operating range. Check unit, scale, estimated nature and whether level is available for the particular machine. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Odometer or distance
What the field means. A distance-related machine value, where applicable, with its own unit and counter scope. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Earth-moving machines do not all have a meaningful road-vehicle odometer; no reading must not become zero distance. Preserve not-supported versus missing statuses and identify whether units are kilometres, miles or another measure. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Status alert
What the field means. A reported message, condition or status field that may indicate attention is needed under vendor definitions. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. An alert is not automatically a verified mechanical fault, shutdown order or permission to continue working. Keep original code, severity definition, time and diagnostic source; route safety interpretation to qualified personnel. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Null or unavailable
What the field means. A field is absent, unsupported, not yet reported or intentionally withheld, depending on the interface contract. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Replacing missing with zero creates false precision, especially for location, fuel or runtime counters. Retain API-specific null semantics and display a distinct user-friendly state instead of assuming measurement. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Sampling interval
What the field means. The spacing or cadence of observations or report updates, which may differ by sensor and provider. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. A one-minute refresh in the user interface does not mean the machine delivered a fresh measurement every minute. Preserve update frequency, observed-at timestamp and communications delay when presented. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Unit of measure
What the field means. The dimension attached to each reported number, including hours, distance, volume, temperature or altitude. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Literal numbers without units may be impossible to compare or may imply the wrong measure across countries. Validate SI and local display conversions with known test values while retaining original data in its source unit. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Provider response time
What the field means. When an API answered a client call, rather than necessarily when the underlying machine event took place. The exact definition should come from the controlled source dataset or provider contract, not from an assumed equivalence with a familiar automotive expression.
Where translation fails. Calling a fresh API response ‘fresh machine data’ confuses communication with observation. Keep event timestamps, collection time and retrieval time separate in both exports and prose. Before release, confirm that the target label leads a fleet manager to calculate or interpret the same thing as the source label.
Fourteen worked fictional fleet-data cases
Case 1. The excavator that appeared to move
Fictional source. Source: ‘Last position 06:20 UTC, lat 1.3000, long 103.8000; dashboard retrieved 10:20 UTC.’
Data interpretation. The server answered four hours after the last underlying position. The source never claimed continuous or current tracking. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Excavator currently at 1.3000, 103.8000 as of 10:20.’ The dashboard timestamp is substituted for the position timestamp. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: retain both times and label last reported location, with update age or unknown present location where supported. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: compare point on map, event-time field, retrieval log and any stale-data indicator; never infer actual machinery movement from the translation. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 2. Lifetime counter mistaken for shift total
Fictional source. Source: ‘Operating hours 5,182.6 h recorded 17:00; shift length eight hours.’
Data interpretation. The first figure is cumulative and the second a separately scheduled period. They cannot be combined without an earlier counter value and definition. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘The excavator operated 5,182.6 hours during the shift.’ This creates an impossible duration. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: show accumulated operating hours as 5,182.6 h and describe the eight-hour shift separately. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: ensure reports do not subtract a guessed baseline or equate engine-on time with productive work. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 3. Fuel used versus fuel remaining
Fictional source. Source: ‘Cumulative fuel used 37,410 L; current fuel level 42%.’
Data interpretation. The numbers represent different dimensions and time scopes; one is volume consumed, the other an estimated proportion remaining. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘37,410 litres left, 42 litres used.’ It reverses both units and concepts. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: retain cumulative used and remaining percentage with corresponding timestamps. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: compare field names, original units, model display and any conversion in the localisation layer. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 4. Machine ID with leading zeroes
Fictional source. Source: ‘Equipment ID 000742-A, PIN 8X7P40001.’
Data interpretation. The equipment ID and PIN are separate references, even if one help-desk screen shows only one. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Equipment No. 742-A’ or ‘VIN 8X7P40001’ when the record says PIN. Leading zeros and identity type vanish. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: preserve both exact strings, and label the second PIN, not VIN, unless the data contract uses that term. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: perform exact-string comparisons in target export and source system. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 5. Missing data converted to zero
Fictional source. Source: ‘IdleHours: null; OperatingHours: 1,148.3.’
Data interpretation. The provider has not supplied idle hours. A zero would be a real measured result with different implications. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘No idling reported: 0 h.’ This fabricates a clean efficiency claim. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: display ‘idle-hours data unavailable’ while preserving the available operating-hours counter. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: compare null, omitted, not applicable, zero and error states in the actual API contract. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 6. 24-hour consumption mistaken for calendar day
Fictional source. Source: ‘Fuel used in preceding 24 hours: 64.0 L at 14:30 UTC.’
Data interpretation. The source interval runs backward from a specified reading time, not necessarily midnight to midnight. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Fuel consumed today: 64 L.’ The calendar boundary is invented. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: keep preceding 24 hours and the 14:30 UTC reference time. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: test daylight-saving shifts and local time conversion without changing the original rolling-window meaning. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 7. Update too old for dispatch
Fictional source. Source: ‘Last known machine location 13:10 yesterday; connectivity lost.’
Data interpretation. The coordinate may remain useful for historical search but not as a verified active location. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Machine standing at site gate now.’ The expression creates false certainty and hides an outage. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: preserve last known and communications loss. Avoid operational recommendations not in the source. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: have dispatch staff state whether data are live, delayed or unknown after reading the translation. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 8. Latitude and longitude inverted
Fictional source. Source: ‘Latitude -23.558, Longitude 46.639; last position logged at 09:40.’
Data interpretation. Coordinate order and signs determine a physical point. Language-specific comma conventions can obscure machine-readability. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: swap the two values or strip the minus sign to make a familiar-looking display. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: label latitude and longitude separately and keep precision and sign unless a documented format transform is applied. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: plot the result on a verified map and compare the raw pair to the displayed coordinate. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 9. A unit conversion with rounding
Fictional source. Source: ‘Distance 12.4 km’ for an applicable machine type.
Data interpretation. The number may be shown in miles for a target-market dashboard, but conversion requires explicit unit and an agreed precision. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: display ‘12.4 miles’ by changing the unit label only. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: retain 12.4 km or apply a validated conversion while marking the displayed value as converted and not altering the source record. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: compare round-trip values and the intended number of decimals, avoiding false precision. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 10. Counter reset after equipment repair
Fictional source. Source: ‘Counter 7,900 h on Monday; replacement instrument reports 14 h on Thursday, reset documented.’
Data interpretation. A lower value after a component change is not necessarily negative operation or proof of data corruption. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Machine lost 7,886 operating hours.’ This derives a misleading negative number. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: keep the documented reset and distinguish new instrument reading from total lifetime work history. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: reconcile maintenance record, counter replacement and downstream analytics rules before writing a utilisation claim. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 11. OEM calls it idle, fleet calls it standby
Fictional source. Source: ‘Provider field classifies mode as standby; fleet dashboard maps same value to idle.’
Data interpretation. The two labels may reflect different operational conventions; one mapping must not be treated as a universal fact across brands. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: merge all standby minutes into certified idle time for comparative efficiency statements. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: preserve source field and flag the mapping as a fleet-level categorisation requiring technical acceptance. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: collect per-OEM definitions and test mixed-fleet comparisons before publishing a benchmark. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 12. Unresolved diagnostic code
Fictional source. Source: ‘Alert code D-19 active; operator note says intermittent sensor warning, repair status unknown.’
Data interpretation. A code and a human observation are not equivalent to a confirmed fault or corrective action. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Severe hydraulic failure repaired.’ Both severity and repair status are invented. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: keep alert code, active status, observation attribution and ‘repair status unknown’. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: confirm proprietary code dictionary with the manufacturer and keep work authorisation outside the translator’s role. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 13. Daily report with two machine identities
Fictional source. Source: ‘Asset tag T-18 reassigned to replacement loader; serial SN-0045 moved to retired list.’
Data interpretation. An organisational asset tag can be reused or reassigned while serial numbers identify different physical machines. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: treat T-18 as proof the new loader has SN-0045, and merge maintenance histories. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: preserve the reassignment event and separate equipment identity from current asset-label assignment. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: validate effective date, serial history and dashboard join keys to avoid reporting maintenance against the wrong machine. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 14. GPS position suppressed for privacy
Fictional source. Source: ‘Location data withheld by customer policy; fuel and operating hours available.’
Data interpretation. Missing location is deliberate governance, not proof that the machine has no position or no GPS device. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘GPS hardware absent.’ The translation invents a technical diagnosis. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: retain policy-based withholding where the source supplies that reason, and do not reveal coordinates from another export. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: confirm recipient permissions and ensure localised tooltips do not disclose restricted machine movement data. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Case 15. Emission number without calculation basis
Fictional source. Source: ‘Fleet summary estimates CO2 from reported fuel using project-defined factor; not directly measured.’
Data interpretation. An estimate based on fuel is not an exhaust sensor measurement or a universally valid certification result. First identify which underlying field the sentence describes, its unit, observation time and provenance. Those are parts of its meaning even if the visible UI compresses them.
Misleading translation. Unsafe: ‘Machine directly measured certified emissions.’ It upgrades both method and assurance level. A superficially clear label is still inaccurate when it causes a user to compare unlike measurements or overstate machine availability.
Defensible target rendering. Better: retain estimated, factor basis and limitation. If the factor is not given, request it rather than supplying a generic default. Keep protected codes and numbers intact while translating the explanatory words into plain, technically accurate language.
QA evidence. Review: preserve methodology reference, fuel window and assurance status in any translated sustainability report. Compare the translated screen with an unchanged source payload and ask a machinery specialist to interpret the output without reading the source language.
Twelve-stage workflow for multilingual fleet reports
1. Define the actual audience
Identify whether the text goes to a site foreman, service engineer, project owner, contractor, data integrator or end-customer. The glossary must support their decision without adding interpretations unsupported by the telemetry record. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
2. Freeze the document and interface version
Record the version of ISO/TS 15143-3, provider API specification, OEM software and dashboard release. A mapping verified for one version may not apply to the next. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
3. Draw the data path
Identify machine source, telematics logger, provider server, customer API and report layer. Mark which timestamps belong to collection, server receipt, query and display. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
4. Protect machine identity
Lock serial, equipment ID, PIN/VIN, manufacturer model and digital record keys. Check leading zeros and characters that could be lost through spreadsheet import. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
5. Create a unit dictionary
For each field, record type, source unit, target display unit, precision, accepted conversion, counter scope and reset rules. Never let translation alter raw numerical values. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
6. Control time and location
Keep UTC/offset semantics, coordinate order, signs, timestamp precision and ‘last known’ language. Show staleness where the UI design allows. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
7. Preserve missing states
Document the difference between null, unsupported, unavailable, withheld, error and a measured zero. Verify that localised labels retain these distinctions. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
8. Separate counters and intervals
Cumulative hours, period hours, cumulative fuel and preceding-24-hour fuel require different labels. Do not compute a difference unless the system specification requests and defines it. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
9. Keep source-specific alerts
Translate generic descriptions carefully while preserving exact alert codes and statuses. Avoid inventing diagnostic certainty, repair approval or safety clearance. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
10. Test realistic edge cases
Include a counter reset, disconnected machine, stale GPS value, negative latitude, mixed units and repeated asset tag. Verify outputs in every target locale. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
11. Obtain bilingual engineering review
Ask an engineer or data integrator to explain the same fleet state using the source and target screens, then compare discrepancies in identity, units, status and time. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
12. Release and monitor changes
Archive approved strings and provider mappings. When API fields, machine types or firmware change, retest the relevant labels rather than assuming a past sign-off covers them. Record the evidence for the decision and the receiving role that will rely on it. A rigorous translation handoff includes not only a polished display phrase but also the source field, approved term and a reproducible test case.
Practice exercises for translation and fleet-data teams
Build a five-machine identity sheet
Invent five machines with asset tags, serial numbers and model codes, including leading zeros and one reassigned tag. Translate the descriptions but not the identifiers. Compare exact strings and check that each history remains attached to the same physical machine. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Audit two kinds of time
Use a fictional record containing event-at, server-received-at and dashboard-refreshed-at. Have students write a user-facing sentence that preserves which time means what. Test one stale reading across a calendar boundary. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Mixed-unit fuel analysis
Create source data containing 100 L cumulative fuel, 35% fuel level and 8.5 L preceding-24-hour use. Ask learners to translate all three without presenting one as another. Require a second reviewer to infer the correct comparison. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Missing-value challenge
Prepare five UI states: measured zero, null, unsupported, access denied and delayed. Write distinct target-language messages and explain what a manager is entitled to conclude from each. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Location precision exercise
Write coordinates with negative and positive signs and a source timestamp. Check a bilingual display in decimal-point and decimal-comma locales, then plot the untouched numeric payload to confirm there was no coordinate shift. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Live versus last-known interview
Show a user interface reporting a location last received yesterday but refreshed a minute ago. Ask the reader where they believe the machine is now. Revise the wording until they distinguish communication freshness from observation freshness. The final exercise record should name the mistaken inference, the protected source field and the minimal translation change that resolves the ambiguity without inventing a new operational fact.
Questions fleet translators regularly ask
Does every ISO/TS 15143-3 machine transmit every defined field?
No. Real provider offerings and machine capabilities vary, and a published schema does not guarantee a live populated value in every response. Verify actual data coverage by provider, machine and software version. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Are operating hours the same as billable job hours?
Not necessarily. A cumulative machine counter may count engine-running or a particular operating definition, while billable job hours follow project and commercial rules. Keep the source labels and do not translate one into the other. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Why is a new server response not a current location?
The server can answer a request using a previously stored machine position. Retain the observation timestamp and last-known language. Query freshness and event freshness are different concepts. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Should serial numbers ever be translated?
No, not as prose. Preserve the original manufacturer and equipment identifiers exactly unless a documented system format transformation is required. Localise the field label, not the identity value. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
How should missing fuel values appear?
Use the receiving system’s approved state such as unavailable, unsupported or not reported, depending on the source. Zero is reserved for actual measured or defined zero, not an absence of data. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Can I convert kilometres to miles or litres to gallons?
A validated display conversion may be appropriate for the locale, but never change the source value by relabelling. Specify units and precision, test round trips and retain raw data for reconciliation. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Is a fleet alert proof that equipment is unsafe?
No. Codes and alert categories require manufacturer-specific and operational interpretation. The translator preserves code, severity as defined, status and description, while qualified personnel make equipment decisions. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Why not rely entirely on machine translation?
Machine translation can assist drafts, but a false friend among engineering fields or a silent unit conversion can change fleet decisions. Use a controlled terminology map and tests against actual API data with technical review. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
Does this guide certify AEMP or ISO interoperability?
No. It teaches language and data meaning using fiction. Actual implementation needs the controlled technical specification, partner API documentation, tests and appropriate conformity or contractual review. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
What does a 2026 draft mean for a 2020 implementation?
A newer draft is under development. Until a successor is formally published and adopted for the relevant contract, do not describe draft provisions as already controlling. Record the applicable approved edition and plan migration separately. In all cases the release test is whether a fleet manager can distinguish a measured value, a calculated value, a stale value and an unknown value from the translated record alone.
References and connected learning
Official scope: ISO/TS 15143-3:2020 — Worksite data exchange: Telematics data. The general method for crossing languages belongs to Master Art of Translation. Vocabulary teaching and learning remain with the protected Vocabulary Learning Hub; English-side grammar, audience and text meaning remain with How English Works. This specialist resource owns translating machinery telematics representations only, not a competing general translation or machinery-information hub.
The central rule can fit on one line in the fleet manager’s checklist: translate the explanation, preserve the machine data and prove that the receiver will draw the same conclusion. A report becomes genuinely multilingual when its meaning, not merely its labels, survives the journey.