# Ads B Surveillance > Use when you must assess an ADS-B Out installation and ADS-B In reception geometry against the DO-260B-style performance categories: map the navigation integrity category (NIC) to its containment radius, map the navigation accuracy category for position (NACp) to its 95-percent accuracy bound, map the source integrity level (SIL) to its per-flight-hour probability bound, select the NIC and NACp category whose bound covers a required containment or accuracy value, and compute the 1090 MHz extended squitter radio line-of-sight range between two altitudes. Produces the containment radius, accuracy bound, integrity probability, chosen categories, and coverage range that gate an ADS-B surveillance assessment. Trigger: ads-b-out, ads-b-in, extended-squitter, nic, nacp, sil, containment-radius, ads-b-range. - Skill: `ashfordeou/ads-b-surveillance` (Agent Skill, multi-file: 3 files) - Install (CLI): `npx skillmds@latest add ashfordeou/ads-b-surveillance` - Raw SKILL.md: https://api.skillmd.com/api/skills/ashfordeou/ads-b-surveillance/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - License: Apache-2.0 - Author: ashfordeOU (https://skillmd.com/u/ashfordeou) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/ashfordeou/ads-b-surveillance --- # ADS-B Surveillance (avionics/surveillance/ads-b-surveillance) Use when the task is assessing an ADS-B Out installation and an ADS-B In reception geometry against the DO-260B-style performance categories: the containment radius guaranteed by a navigation integrity category (NIC), the 95-percent horizontal accuracy bound of a navigation accuracy category for position (NACp), the per-flight-hour integrity probability of a source integrity level (SIL), the NIC and NACp whose bounds cover a required containment or accuracy value, and the 1090 MHz extended squitter radio line-of-sight range between two altitudes. This leaf is the second of the avionics/surveillance pack; avionics/surveillance/tcas-resolution-advisory owns the TCAS sense logic on measured state, while this leaf sizes the surveillance performance claims that gate an equipage assessment. The category bounds are paraphrased DO-260B summary values held as module constants in the logic file, never a reproduction of the MOPS tables. Pure Python stdlib, deterministic and offline. Units: containment radius and accuracy in metres, altitude in feet, range in kilometres, integrity as a probability per flight hour. ## Domain quick reference - NIC containment radius: each navigation integrity category 1 to 11 maps to the containment radius the position source guarantees, from 7.5 m at NIC 11 down to 37040 m at NIC 1; NIC 0 means unknown and maps to None. NIC 8, for example, bounds containment to 185.2 m (one tenth of a nautical mile). Exact per-category values live in the NIC_RADIUS_M module constant in ads_b_surveillance_logic.py. - NACp accuracy: each navigation accuracy category for position 1 to 11 maps to the 95-percent horizontal accuracy bound, from 3 m at NACp 11 down to 18520 m at NACp 1; NACp 0 is unknown. NACp 9 bounds the 95-percent position error to 30 m. Values live in NACp_95_M. - SIL integrity: each source integrity level 1 to 3 maps to the maximum probability of an undetected failure per flight hour, 1e-7 at SIL 3, 1e-5 at SIL 2 and 1e-3 at SIL 1; SIL 0 is unknown. Values live in SIL_PROB. - Category selection: nic_for_radius(required_radius_m) and nacp_for_accuracy(required_95_m) walk the categories from the tightest bound (highest number) down to category 1 and return the first whose bound is >= the required value, the least-integrity category that still bounds the requirement. A returned 0 means no category covers the requirement, not even category 1. - Radio line-of-sight range: d = RANGE_COEFF * (sqrt(h_own * FT_TO_M) + sqrt(h_other * FT_TO_M)) with RANGE_COEFF = 4.12 km per sqrt(metre) and FT_TO_M = 0.3048, the standard-atmosphere 4/3-earth radio horizon used for 1090 MHz extended squitter coverage. - Assessment bundle: adsb_assessment(nic, nacp, sil, alt_ft_own, alt_ft_other) returns {containment_radius_m, accuracy_95_m, integrity_prob, range_km} in one call. ## Workflow 1. Fix the assessment state: the candidate NIC, NACp and SIL from the position source and the own and other altitudes in feet. 2. Map the categories to bounds with nic_containment_radius(nic), nacp_accuracy(nacp) and sil_probability(sil); category 0 returns None, meaning the value is unknown. 3. For a required containment radius or 95-percent accuracy, select the covering category with nic_for_radius(required_radius_m) or nacp_for_accuracy(required_95_m); a returned 0 flags that no category bound covers the requirement. 4. Compute the reception coverage with adsb_range_km(alt_ft_own, alt_ft_other), passing 0 ft for the altitude of a ground ADS-B receiver. 5. Bundle the whole case with adsb_assessment(nic, nacp, sil, alt_ft_own, alt_ft_other) and read containment_radius_m, accuracy_95_m, integrity_prob and range_km. 6. Confirm the deterministic checks with the contract test scripts/test_ads_b_surveillance.py. ## Worked example Own ship at 10 000 ft receiving an ADS-B In target at 30 000 ft, with the source reporting NIC 8, NACp 9, SIL 2. - nic_containment_radius(8) = 185.2 m: the position source guarantees containment within 185.2 m. - nacp_accuracy(9) = 30.0 m: the 95-percent horizontal position error stays within 30 m. - sil_probability(2) = 1e-5: at most one undetected failure per 100 000 flight hours. - nic_for_radius(100.0) = 8: NIC 9 bounds only 75 m, too small for the 100 m required containment radius; NIC 8 bounds 185.2 m and covers it, so NIC 8 is the chosen category. - nacp_for_accuracy(50.0) = 8: NACp 9 bounds only 30 m, too small for the 50 m required accuracy; NACp 8 bounds 92.6 m and covers it. - adsb_range_km(10000, 30000) = 621.4 km: sqrt(3048) = 55.21, sqrt(9144) = 95.62, sum 150.83 times 4.12 gives 621.4 km, inside the 600 to 650 km band for the altitude pair. - adsb_assessment(8, 9, 2, 10000, 30000) returns {containment_radius_m: 185.2, accuracy_95_m: 30.0, integrity_prob: 1e-5, range_km: 621.4}. ## Verification - nic_containment_radius(8) is 185.2 m and nacp_accuracy(9) is 30.0 m, exactly the table entries (equal within 1e-9 relative). - nic_for_radius(7.5) is 11 and nic_for_radius(1e6) is 0, the exact-bound and no-coverage ends of the selection rule; nacp_for_accuracy(50.0) is 8. - adsb_range_km(0, 0) is 0.0 and the 10 000/30 000 ft case sits inside the 600 to 650 km bound; SIL 3 gives the stricter probability (1e-7 below the 1e-5 of SIL 2). - Every out-of-range category (NIC or NACp outside 0 to 11, SIL outside 0 to 3), every negative altitude and every non-positive required radius or accuracy raises ValueError. - Run the contract test offline: python3 scripts/test_ads_b_surveillance.py (34 tests, deterministic). ## Related leaves - avionics/surveillance/tcas-resolution-advisory: the sibling leaf in the surveillance pack; it owns the TCAS II sense logic on measured range and altitude state, this leaf owns the surveillance performance categories. - gnc-autonomy/navigation/gnss-pseudorange-positioning: the position solution whose accuracy and integrity feed the NACp and NIC claims. - gnc-autonomy/navigation/gnss-raim-fde: fault detection that backs a source integrity level claim. - avionics/flight-management/radio-navigation-aids: navaid sensing context for the airborne navigation suite. ## Pitfalls - Treating NIC, NACp, and SIL as one quality number: NIC bounds containment radius, NACp bounds the 95-percent horizontal position error, and SIL is an integrity probability per flight hour - NIC 8 at 185.2 m and NACp 9 at 30.0 m describe different guarantees, and neither is comparable with a SIL 2 probability of 1e-5. - Choosing a category that does not cover the requirement: the selection walks from the tightest bound down and returns the first category whose bound is >= the required value - for a 100 m containment radius NIC 9 (75 m) is too small and NIC 8 (185.2 m) is chosen, so never pick the category whose bound merely comes closest from below. - Reading the selection result 0 as the unknown category: a returned 0 from nic_for_radius or nacp_for_accuracy means no category bound covers the requirement at all (not even category 1's 37040 m or 18520 m), while an input category of 0 maps to None for an unknown value - the two zeros are different verdicts. - Forgetting the feet-to-metres conversion in the range law: the radio horizon uses sqrt(h_ft * FT_TO_M), so 10 000 ft enters as sqrt(3048) = 55.21 and the 10 000/30 000 ft pair reaches 621.4 km - feeding feet straight into the square root without the 0.3048 factor inflates the coverage estimate. - Assigning coverage without altitudes: a zero altitude (a ground ADS-B receiver at 0 ft) contributes zero horizon and the pair range is 0.0 km - the own/other altitude pair is part of the input, and negative altitudes raise ValueError. - Quoting the category bounds as MOPS tables: the NIC/NACp/SIL values are paraphrased summary data in the module constants used for equipage assessment, stated as module data and never as reproduced rtca-do-260b tables - confirm against the current MOPS before a formal claim. ## Behavior contract (gate 3) Run the deterministic contract test (stdlib unittest, offline): python3 scripts/test_ads_b_surveillance.py The test covers the exact table anchors (NIC 8 at 185.2 m, NACp 9 at 30.0 m, SIL 2 at 1e-5), table spot checks with unknown-category None returns, the selection rule for required containment radii and accuracies including the exact-bound and no-coverage cases, the 10 000/30 000 ft range worked example (621.4 km inside the 600 to 650 km bound), the zero-altitude range, the assessment bundle, ValueError rejection of out-of-range categories, negative altitudes and non-positive requirements, and determinism of the repeated calls. ## Compliance - Standards referenced, not reproduced: rtca-do-260b (DO-260B) is a gated RTCA standard; the category bounds above are paraphrased summary values used in equipage assessments, stated as module data, never as reproduced MOPS tables. - compliance: STANDARDS-REF, gated: false.