Partner with usTalk to fleet team

Technology

Every vehicle, every trip, every rupee.Instrumented.

Bluwheelz is a software company that runs vehicles: telemetry every three minutes, exceptions every fifteen, and one audited record per vehicle from sourcing to exit.

  • 3-minute telemetry poll
  • 15-minute exception cycle
  • One relational record
  • 20 audited lifecycle steps
6 layers, one record per vehicle from OEM sourcing to exit.

The stack

Six layers. The second one is the difference.

Capture is common. Dashboards are common. Correcting a raw feed before anything reads it is not.

5 live1 planned

  1. CaptureEverything enters where the work happens, not afterwards.Live
  2. Ingest and normaliseThe differenceRaw feeds corrected before anything downstream reads them.Live
  3. Unified data coreOne relational system of record, not four joined after the fact.Live
  4. IntelligenceRules first, models second, both writing to the same record.Live
  5. InterfacesSix audiences, one record underneath each of them.Live
  6. LedgerPlannedDesigned to anchor the audit trail. Not in operation.

02What normalisation fixes

  • Coordinate and speed bounds: impossible positions rejected at the door
  • Odometer repair: resets, rollovers and stuck counters reconciled
  • Canonical client resolution: duplicate spellings collapse to one entity
  • Every correction recorded, never silently overwritten

Vehicle lifecycle

Twenty steps from OEM sourcing to exit, each one audited.

Each step names its system, what it captures and what the audit keeps.

AI workflows

The model proposes. A person decides.

Every workflow ends with a named human decision, written back against the recommendation that prompted it.

  1. SignalComputed from the operating record.
  2. Model proposesA slot, a score or a cluster, with its reasons.
  3. Person decidesAnd the decision is written back against it.

In buildAll three workflows

  • Preventive scheduling

    Signal
    Distance, elapsed time and duty intensity cross a threshold.
    Model proposes
    Ranks vehicles due for service and proposes a slot that does not collide with a committed shift.
    Person decides
    The workshop lead confirms or moves the slot. Nothing is booked automatically.

  • Predictive health

    Signal
    Degradation trend, charging behaviour, temperature and fault history.
    Model proposes
    Produces a health score and a ranked watch list, with the factors that drove it.
    Person decides
    Maintenance decides what is actioned, and the decision is recorded against the score.

  • Exception intelligence

    Signal
    The 15-minute exception log, read across weeks rather than shifts.
    Model proposes
    Clusters exceptions to separate one-off events from recurring causes.
    Person decides
    City ops decides what changes: a route, a roster, a charging window.

Ledger

Planned

Designed to anchor the audit trail. The one that already exists.

Planned, not in operation. A ledger would let a client verify a report without being handed the underlying record.

What it would add

  • Immutable MIS anchoring

    A revised figure would show the revision.

  • Carbon credit MRV

    Traceable to individual trips, not an annual estimate.

  • Client-verifiable reporting

    A client could check the report against what was recorded.

Recorded today

Anchor
  1. Onboarding auditStep 02
    Planned
  2. Pre-trip checkStep 10
    Planned
  3. ExceptionStep 13
    Planned
  4. Charge sessionStep 15
    Planned
  5. Job cardStep 19
    Planned
  6. RenewalStep 20
    Planned

Every entry is written by the workflow today. Anchoring would add verification, not new data.

Carbon credit issuance in India operates under an evolving regulatory framework. Nothing here describes credits as issued, sold, traded or certified.

Data security

Principles, not a topology diagram.

We publish how access is governed, not how it is built. The detail is for a technical reviewer under NDA.

  • Role-based access01
  • Least privilege by default02
  • Encryption in transit and at rest03
  • Access audit logging04
  • Data minimisation05
  • Defined retention06
  • DPDP Act 2023 alignment07

Next step

Bring your engineers. We will open the system.

The data model, the exception engine and the audit trail, walked through by the people who built them.

Request a technical walkthroughTalk to fleet team

We reply within one working day.

  1. 01Data model and normalisation layer

  2. 02Exception engine and KPI computation

  3. 03Audit trail and access governance