Configuration and control

Example 1 Checking the parameter file against requirements

The team’s requirements: for a 6S battery, the low level at 3.5 V per cell and the critical level at 3.3 V per cell, and both levels must command more than a warning; RC and GCS failsafes must be on; the fence must be on with a ceiling of no more than 90 m; and the RTL height must be at least 5 m above the tallest trees around the field, which are 25 m.

params = {   # read from the flight controller's .param file (hypothetical values)
    "BATT_LOW_VOLT": 21.0, "BATT_CRT_VOLT": 19.8, "BATT_FS_LOW_ACT": 2, "BATT_FS_CRT_ACT": 1,
    "FS_THR_ENABLE": 1, "FS_GCS_ENABLE": 0, "FENCE_ENABLE": 1, "FENCE_ALT_MAX": 100,
    "RTL_ALT": 2500,
}
CELLS, TREE_M = 6, 25
rules = [   # requirement, check function
    ("low voltage >= 3.5 V/cell", lambda p: p["BATT_LOW_VOLT"] >= 3.5 * CELLS),
    ("critical voltage >= 3.3 V/cell", lambda p: p["BATT_CRT_VOLT"] >= 3.3 * CELLS),
    ("battery actions not 'warn only'", lambda p: p["BATT_FS_LOW_ACT"] > 0 and p["BATT_FS_CRT_ACT"] > 0),
    ("RC failsafe enabled", lambda p: p["FS_THR_ENABLE"] != 0),
    ("GCS failsafe enabled", lambda p: p["FS_GCS_ENABLE"] != 0),
    ("fence on, ceiling <= 90 m", lambda p: p["FENCE_ENABLE"] == 1 and p["FENCE_ALT_MAX"] <= 90),
    ("RTL altitude >= trees + 5 m (RTL_ALT in cm)", lambda p: p["RTL_ALT"] / 100 >= TREE_M + 5),
]
failed = 0
for text, check in rules:
    ok = check(params)
    failed += not ok
    print(f"{'PASS' if ok else 'FAIL'}  {text}")
print(f"{failed} rule(s) failed -> {'do not fly' if failed else 'cleared for pre-flight checks'}")

PASS  low voltage >= 3.5 V/cell
PASS  critical voltage >= 3.3 V/cell
PASS  battery actions not 'warn only'
PASS  RC failsafe enabled
FAIL  GCS failsafe enabled
FAIL  fence on, ceiling <= 90 m
FAIL  RTL altitude >= trees + 5 m (RTL_ALT in cm)
3 rule(s) failed -> do not fly

The check finds the GCS failsafe off, the fence ceiling above the limit, and an RTL height of 2500 cm, or 25 m, exactly at treetop height. Anyone reading 2500 as metres would think it very safe. Unit mistakes like this are why a program should do the checking instead of the eye.

Example 2 Recording a baseline and detecting changes

import hashlib

def fingerprint(text):
    return hashlib.sha256(text.encode("utf-8")).hexdigest()[:16]   # show the first 16 characters for readability

baseline = {
    "params": "BATT_LOW_VOLT,21.0\nFS_GCS_ENABLE,1\nRTL_ALT,3000\n",
    "harness": "W02 BEC A -> companion 5V 22 AWG\n",
    "payload": "thermal camera, 4 soft dampers\n",
}
today = dict(baseline, params="BATT_LOW_VOLT,21.0\nFS_GCS_ENABLE,1\nRTL_ALT,300\n")
record = {k: fingerprint(v) for k, v in baseline.items()}
for item, text in today.items():
    now = fingerprint(text)
    status = "unchanged" if now == record[item] else "CHANGED since baseline"
    print(f"{item:<8} {now}  {status}")

params   4d64c9c54070644e  CHANGED since baseline
harness  58b531d879b843c2  unchanged
payload  f7aa99efe1967fc3  unchanged

Mistyping RTL_ALT with one zero missing changes the parameter file’s hash at once, even though someone scanning a list hundreds of lines long would not see it. The team must find the cause, fix it, and record a new baseline through a change request as in UAT 494.