Telemetry Platform Roadmap
A structured implementation plan for the full telemetry stack — from CSV import through to live streaming, track mapping, and advanced analysis at scale.
Architecture Principles
Platform Design Philosophy
Single Canonical Schema
Both CSV import and live telemetry streaming must map into one internal schema. The ingestion source is irrelevant — the data model is consistent.
Track Mapping as a Subsystem
The track map engine is independent. It consumes positional data and outputs a reusable TrackModel that all sessions, analyses, and live views reference.
Raw vs Derived Storage
Raw telemetry samples are immutable and append-only. Derived analytics (lap stats, corner times, consistency scores) are computed separately and are rebuildable.
Import-First → Live Cloud Evolution
The platform starts with CSV-driven import workflows. The architecture is designed so this evolves cleanly into direct server-side session writes without breaking existing data.
Implementation Stages
Objective
Define the canonical telemetry schema that every other system in the platform depends on. This is the single source of truth for all data — past, present, and live.
Why This Matters
Every downstream system (CSV import, live streaming, track mapping, analysis, storage) must agree on the same structure. Defining this first prevents costly rework and ensures that data from any source (CSV or live) flows into one consistent model.
Implementation Tasks
- Define canonical telemetry sample schema (fields, types, units)
- Define Session, Lap, Sample, and TrackModel entity structures
- Establish naming conventions (snake_case, SI units, channel IDs)
- Classify channels as required vs optional vs derived
- Define the lap boundary contract (start/end conditions)
- Document the schema as the platform's internal API contract
Dependencies
- None — this is the starting point
Risks & Challenges
- Changing the schema later is expensive — take time to design it well
- Over-engineering too early can slow down iteration
Expected Output
A locked, versioned data schema document. All future work references this.
Stage Index
Recommended Build Order
Foundation / Data Contract
Everything else depends on a locked schema.
CSV Import Pipeline
Populate real data, validate the schema in practice.
Storage Architecture
Establish the two-tier model before live data arrives.
Track Map & Alignment Engine
Required for spatial analysis and live position mapping.
Near-Live Telemetry
Validate ingestion and session state at low risk.
True Live Streaming
Build on proven ingestion path with full WebSocket flow.
Advanced Analysis & Scaling
Layer intelligence on top of a solid data foundation.