IdeaWell

Project ideas from Hacker News discussions.

NTSB issues investigative update on B-767 runway excursion accident in Miami

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

Theme 1: Corporate / operational pressure influencing go‑around decisions
Commenters frequently suggested that airlines or cargo operators (especially Amazon‑linked 21 Air) create implicit pressure to avoid go‑arounds to keep schedules or fuel costs low.
- wanderingmoose: “I think the real question is why didn't the pilots do a go around. I bet there are underlying pressures from the airlines and/or amazon to not do a go around to avoid schedule delays or fuel costs.”
- pseudosavant: “Could a company that won't let delivery drivers stop to pee possibly be pushing their pilots that hard? Yes. Yes, they could.”
- ElProlactin (citing a lawsuit): “Seuring's attorney … told the judge that another 21 Air pilot … 'was being pressured to fly in periods of time that were too short to comply with the proper amounts of rest, and that implicated safety.'”

Theme 2: Pilot decision‑making / human factors (get‑there‑itis, situational awareness, airmanship)
Many saw the crash as a failure of judgment, with pilots persisting on an unstable approach despite cues to go around.
- psunavy03: “The problem occurred way behind that, when they refused to take a botched approach around. Everything else was just them doubling down on a bad decision over and over.”
- MBCook: “Get‑there‑itis is absolutely a thing. No big airline would be dumb enough to say it that stuff explicitly but the pressures still exist.”
- loeg: “They were in a bad situation long before they touched the ground and should have gone around even earlier.”

Theme 3: Technical details of the landing/go‑around sequence (spoilers, weight‑on‑wheel, engine spool‑up)
A substantial thread focused on why the aircraft’s automatic systems didn’t activate and the mechanics of a late go‑around.
- buildsjets: “The Boeing 767 does not have a Flaps 40 setting. The 'Forty' callout was an automated annunciation of the radar‑altimeter measured altitude above the runway.”
- ofcrpls: “The lack of contact of the left wheels on the ground, which the Weight of Wheel sensor deployment depends on… worsened the whole situation as the crew seems to be unaware why they weren't losing speed after they'd 'landed'.”
- xenadu02: “No, the left main was not on the ground after touchdown for some time. The aircraft will not auto‑deploy the spoilers nor autobrake until it senses weight on wheels.”

🚀 Project Ideas

Generating project ideas…

Approach Stability Monitor & Go-Around Advisor

Summary

  • Real-time AI-driven avionics add-on that monitors speed, altitude, glidepath, flap/gear configuration, and weight-on-wheel sensors to detect unstable approach conditions and issues audible/visual go-around recommendations before it's too late.
  • Core value: Prevents landing accidents by providing timely, objective decision support that counters get‑there‑itis and cockpit hierarchy issues.

Details

Key Value
Target Audience Commercial cargo and passenger airlines, flight departments, pilot unions
Core Feature Real-time approach stability analysis with go‑around advisory alerts
Tech Stack C/C++ for embedded avionics, Rust for safety‑critical logic, Python for ML models, CAN bus/ARINC 429 interfaces, optional Android/iOS tablet companion
Difficulty High
Monetization Revenue-ready: SaaS licensing per aircraft + optional hardware kit

Notes

  • HN commenters noted “11 mandatory go‑around triggers were ignored” (toomuchtodo) and that “the pilots should have gone around at 1000 feet” (JumpCrisscross). This tool would surface those triggers automatically.
  • Provides concrete utility for safety teams and could reduce accidents caused by delayed go‑arounds, a frequent discussion point on HN.

Crew Resource Management (CRM) Training Simulator with AI Coach

Summary

  • Immersive VR/desktop simulator that recreates high‑stress approach scenarios (like the Miami cargo crash) and trains pilots and first officers on assertive communication, situational awareness, and go‑around decision making.
  • Core value: Improves CRM skills and reduces hierarchy‑related failures by offering repeatable, data‑driven practice with an AI coach that gives feedback on verbal cues and adherence to SOPs.

Details

Key Value
Target Audience Flight training departments, airlines, pilot unions, aviation academies
Core Feature Scenario‑based CRM training with real‑time AI feedback on communication and decision making
- Tech Stack Unity or Unreal Engine for VR, Python/TensorFlow for speech‑act analysis, AWS/GCP for cloud scoring, optional eye‑tracking hardware
Difficulty Medium
Monetization Revenue-ready: Per‑pilot subscription or license fee per simulator instance

Notes

  • Commenters highlighted the first officer’s reluctance to challenge the captain (drillsteps5: “PM finally said 'too fast, go around' but PF ignored him”) and the need for better interpersonal dynamics (outworlder: “lack of adequate FAA mandated training around interpersonal dynamics”). This simulator directly trains those skills.
  • Airlines and safety regulators have shown interest in CRM upgrades; the tool would generate discussion on HN about human factors and training efficacy.

Anonymous Safety Reporting & Pressure Detection Platform

Summary

  • Secure, anonymized web‑based platform where pilots can report corporate pressure, unsafe scheduling, intimidation, or observed safety deviations without fear of reprisal; aggregates reports to detect patterns (e.g., repeated pressure to avoid go‑arounds) and generates actionable safety insights for airlines and regulators.
  • Core value: Encourages a just culture by protecting reporters while providing data‑driven evidence of systemic pressures that contribute to accidents.

Details

Key Value
Target Audience Pilots (cargo & commercial), airline safety departments, aviation regulators, pilot unions
Core Feature Anonymous reporting with encrypted submission, pattern detection algorithms, dashboard for safety trends
Tech Stack React/Frontend, Node.js/Express backend, PostgreSQL database, end‑to‑end encryption (libsodium), optional Tor hidden service for extra anonymity
Difficulty Medium
Monetization Revenue-ready: Subscription fees for airlines/regulators to access aggregated analytics; free for individual reporters

Notes

  • Multiple HN users speculated about Amazon/airline pressure to avoid go‑arounds (wanderingmoose, pseudosavant, ElProlactin) and wished for evidence; this platform would give them a way to surface such concerns safely.
  • The NTSB’s lack of “commercial pressure” hits (dghlsakjg) shows a gap this tool could fill, sparking discussion about data transparency in aviation safety.

Go‑Around Decision Support & Post‑Flight Analytics

Summary

  • Service that automatically processes flight data recorder (FDR) and cockpit voice recorder (CVR) streams (or simulated data) to detect missed go‑around opportunities, quantify deviations from stable approach criteria, and produce debrief reports for operators.
  • Core value: Turns raw flight data into clear, actionable insights that help pilots learn from near‑misses and airlines improve SOPs without manual investigation.

Details

Key Value
Target Audience Airlines, flight safety operators, MROs, accident investigation bureaus
Core Feature Automated detection of ignored go‑around triggers and post‑flight analytics reporting
Tech Stack Python/pandas for signal processing, Apache Kafka for data ingestion, ML models (scikit‑learn/TensorFlow) for anomaly detection, Grafana/PowerBI for visualization, secure cloud storage (AWS S3)
Difficulty High
Monetization Revenue-ready: Per‑flight‑hour pricing or annual contract per fleet

Notes

  • Commenters loved the idea of “11 mandatory go‑around triggers were ignored” (toomuchtodo) and wanted an automated way to capture such events (JumpCrisscross: “They should have gone around at 1000 feet”). This service would flag those triggers in near‑real time and archive them for review.
  • Provides practical utility for safety teams and would likely generate HN debate about the balance between automation and pilot authority, mirroring past discussions on AI in the cockpit.

Read Later