Business Requirements¶
Project: football-trackers — private, real-time youth football tracking. Date: 2026-06-14.
1. Background¶
A youth football coach needs to monitor ~10 players live during training and matches. Off-the-shelf trackers either can't be read from outside (closed BLE, store-and-sync) or are priced on per-player subscriptions unsuited to youth budgets, and the real-time ones are all closed. See market-analysis.md. The decision is to build a custom system.
2. Business objectives¶
| # | Objective |
|---|---|
| OBJ-1 | Give the coach live (sub-second) visibility of every player during a session. |
| OBJ-2 | Own the data end-to-end — local storage, full export, no vendor lock-in. |
| OBJ-3 | Equip ~10 players at a one-time low-hundreds-of-EUR cost with no subscription. |
| OBJ-4 | Reach ~80% feature parity with Catapult/STATSports core physical metrics. |
| OBJ-5 | Be extensible to football-specific (technical) metrics later. |
3. Stakeholders¶
| Stakeholder | Interest |
|---|---|
| Coach (primary user) | Live tablet view, per-player metrics, session management |
| Players / parents (later) | Personal stats, trends, records, engagement |
| Owner / developer | Full-stack build; owns firmware, backend, UI, data |
4. Scope¶
In scope¶
- Per-player WiFi wearable (GPS, later IMU) streaming live telemetry.
- Field network (one outdoor AP) + MQTT broker + owned backend ingest.
- Coach live view: player dots on a calibrated pitch; core physical metrics.
- Local persistence and export.
Out of scope (initially) / not feasible¶
- Reading or integrating any commercial tracker (closed BLE / no consumer API).
- BLE-to-phone as the field transport (range can't cover a pitch).
- Possession %, pass networks, team shape from wearables alone (need camera CV).
- RTK centimetre accuracy (needs a base station; overkill for youth).
- Building a smart ball (can't DIY; tracks the ball, not players).
5. Constraints¶
| # | Constraint |
|---|---|
| CON-1 | Transport must be WiFi/4G, not BLE — BLE range 10–30 m vs a 100×64 m pitch. |
| CON-2 | Real-time requires streaming firmware + a live WiFi link during the session. |
| CON-3 | Consumer GPS accuracy is ~2–5 m — fine for youth trends, not elite-grade. |
| CON-4 | Sourced in Serbia (local suppliers + AliExpress/TME); prices in RSD, FX/customs apply. |
| CON-5 | Solo build — firmware and on-field durability are the real effort, not the backend. |
6. Assumptions¶
- ~10 players per team; ~90-minute sessions.
- One outdoor 2.4 GHz AP covers the pitch (line-of-sight ~100 m).
- The coach has a tablet on the touchline on the same network.
- The owner has an AI/ML-capable desktop (RTX 3060 12 GB) for any future computer-vision work.
7. Guardrail¶
Prove one device end-to-end before buying ten — don't scale hardware until firmware → MQTT → backend → live dot works on a single unit.
8. Key decisions¶
Recorded as ADRs in decisions/: build vs buy, WiFi vs BLE, real-time vs store-and-sync, wearable hardware baseline, technical-metrics sensor strategy.