When Software Determines Human Life — On Safety-Critical Systems
summary
Overview of developing safety-critical software for domains like cars, planes, trains, spacecraft, medical devices and nuclear plants. Covers how safety-critical development differs from standard web/mobile/desktop work, the role of detailed standards and norms (including practices used at organizations like NASA), and analyses of real faults that caused deaths or equipment loss to illustrate safer engineering practices.
AI-written summary · about
chapters (14)
- 0:00Introduction to critical systems
- 3:02Therac-25 radiotherapy error
- 9:30Ariane 5 rocket failure
- 13:28Software error patterns
- 14:32Safety norms and V-model
- 22:43Safety levels and techniques
- 25:40Code and language recommendations
- 31:23Safety from project start
- 33:10Risk analysis and matrix
- 37:04Emergency procedures and redundancy
- 41:16Random errors and cosmic radiation
- 43:45Runtime tests and protections
- 45:29Using existing code and costs
- 47:56Are norms really necessary?
related talks
50mWhen Human Life Depends on Software - Introduction to Safety-Critical Systems
58mSafety-critical systems from the inside
42mDecoding Safe(ty) Linux Architectural Approaches for Critical Systems
32mSoftware Craftsmanship
46mA Systems Approach to Safety and Cybersecurity
42mwhy software development is an engineering discipline
58mIs Software Engineering Still an Oxymoron?
54mArchitecture & Responsible Technology