MB5 — Recovery Performance Instrument (Level 5: Judgement & Responsibility)
Mission Goal
Design an instrumentation plan for a “recovery” scenario (landing, retrieval, rescue, or post-mission inspection). You must handle ambiguity, define what “safe recovery” means, and justify your decisions.
Why This Matters
In professional missions, recovery and post-event analysis are where responsibility shows up: safety, evidence, chain-of-custody, and honest reporting. Level 5 is about judgement under imperfect conditions.
What Data You Collect
- Impact detection (acceleration spikes)
- Orientation change (tilt)
- Environmental conditions during recovery (temp/light)
- Recovery timeline markers (button-press events)
- Integrity flags (e.g., “sensor saturated”, “battery low”)
Hardware / Software Needed
- 1–2 × micro:bit (payload + optional “recovery operator” receiver)
- MakeCode / MicroPython
- Optional: buzzer or LED indicator (if available) for “recover me” mode
- Optional: radio relay from MB4 design
Inputs From Other Teams
- Command & Control: Define recovery protocol (who decides stop/go, what’s acceptable risk).
- Risk & Safety: Identify hazards (drops, cables, crowding) and mitigation steps.
- Data: Define evidence integrity: timestamps, run IDs, and how logs are stored without editing.
- Comms: Define “lost payload” mode (beacon + last known status).
What You Must Produce (Deliverables)
- A written Recovery Instrumentation Plan (what you measure + why).
- A working instrument program that captures recovery-critical events.
- A short incident-style report from a test run, including uncertainty and limitations.
Step-by-Step Build
- Define a recovery scenario (choose one):
- “Payload dropped from 1m onto padded surface; measure impact and orientation change.”
- “Payload moved through a course; detect rough handling and ‘safe transport’ compliance.”
- Define “safe recovery” metrics (thresholds and reasoning).
- Implement event logging:
- Impact event when magnitude exceeds threshold
- Tilt event when angle changes beyond threshold
- Operator markers via button A/B
- Implement integrity flags (battery low, sensor saturation).
- Run a controlled recovery test and log results.
- Write a recovery report:
- What happened (based on data)
- What you cannot be sure about
- What you would change next time
Data Format / Output
run_id,t_ms,event,ax,ay,az,mag,temp_c,light,flags- Events:
IMPACT,TILT,MARK_A,MARK_B,STATUS
Analysis Ideas
- Count impacts and compare against “safe” thresholds.
- Plot orientation change timeline during transport.
- Decide if the payload should be “cleared” or “quarantined” based on data quality flags.
Success Criteria
- Team defines recovery success in measurable terms.
- Instrumentation captures events clearly and consistently.
- Report demonstrates professional judgement: honest limits, safety awareness, and evidence integrity.
Evidence Checklist
- Recovery Instrumentation Plan (1 page)
- Log file / serial output screenshot from test
- Short recovery report (what happened + what you’d improve)
- Photo of test scenario + safety setup
Safety & Privacy
- Use padded drop zones and controlled movement; no throwing.
- Keep walkways clear; assign a spotter if moving around.
- No filming faces; if recording evidence, focus on device + hands only.
Common Failure Modes
- No clear definition of “safe recovery” → data has no decision value.
- Thresholds chosen without justification.
- Logs edited after the fact (breaks evidence integrity).
- Over-claiming certainty from limited sensors.
Stretch Goals
- Add “lost mode”: loud/bright beacon + last status packet.
- Create a decision tree: “If IMPACT > X then quarantine.”
- Design a handover checklist (operator A to operator B) with data continuity.
Scaffolding Example (optional)
You are allowed to reuse structures and formats from other teams — but not their decisions.
Structure: “Mission Data Product” (copy this layout)
- What did we measure?
- How did we measure it?
- What did we learn?
- How confident are we (limits/errors)?
- What would we improve next time?
Example confidence language
- “We trust the trend, not the exact value, because …”