WHY three standards instead of one? Because the failure consequences, the operating environment, and the certification authorities differ. An aircraft is certified by the FAA/EASA per flight; a car is mass-produced for the public; an industrial plant runs continuously. IEC 61508 is the generic ancestor; ISO 26262 is its automotive child. DO-178C evolved separately in aviation but shares the same safety logic.
What does the "C" in DO-178C MC/DC coverage requirement guarantee that decision coverage does not?
That each condition independently affects the decision outcome (others held fixed).
DO-178C: which DAL is the most safety-critical?
DAL A (catastrophic failure condition).
ISO 26262: which ASIL is the most safety-critical?
ASIL D.
IEC 61508: which SIL is the most demanding?
SIL 4.
Simplified 1oo1 average PFD formula?
PFDavg=λDUT1/2.
Why the factor 1/2 in PFD_avg?
It's the average of risk that grows linearly from 0 (just tested) to full over the proof-test interval.
What three parameters determine ASIL in ISO 26262?
Severity (S), Exposure (E), Controllability (C).
Which standard is the generic "parent" of the others?
IEC 61508.
How many objectives does DO-178C define?
71 objectives (number satisfied with independence rises with DAL).
For a decision A && B, how many tests does MC/DC require?
n+1 = 3.
What does "with independence" mean in DO-178C?
Verification done by a different person/team than the developer.
SIL 3 low-demand PFD band?
10−4≤PFD<10−3 (RRF 1000–10000).
What is ASIL QM?
Quality Management — no safety requirement beyond normal good practice.
PFD vs PFH — when is each used?
PFD for low-demand mode (rare actuation); PFH for high/continuous-demand mode.
Recall Feynman: explain to a 12-year-old
Imagine you're building a fence. For your backyard, a wobbly fence is fine. For the lion enclosure at the zoo, you need to prove every post is solid — because if it fails, someone gets eaten. These standards are rulebooks that say: first ask "how bad if this breaks?" If it's "a lion escapes" (catastrophic), you must test super carefully, have a second person double-check everything, and keep records proving you did. If it's "the gate squeaks" (no effect), you can relax. The labels (DAL-A, SIL-4, ASIL-D) are just badges saying "this is the lion-cage level — be extra careful."
Dekho, in teen standards ka ek hi core idea hai: jitna khatarnaak failure, utni zyada rigor. Agar software fail hone se plane gir sakta hai ya car ka brake fail ho sakta hai, toh sirf "test karke ship kar do" nahi chalega — aapko ek disciplined, audit-able process follow karna padta hai aur uska evidence dena padta hai. DO-178C aviation ke liye hai, IEC 61508 generic parent standard hai (sabka baap), aur ISO 26262 uska automotive version hai.
Har standard ek level deta hai jo rigor decide karta hai. DO-178C mein ye DAL hai (A se E, jismein A sabse deadly). IEC 61508 mein SIL hai (1 se 4, jismein 4 sabse deadly). ISO 26262 mein ASIL hai (A se D plus QM, jismein D sabse deadly). Yaad rakhna ye ulta-pulta lagta hai — DO-178C mein A khatarnaak, ISO mein D khatarnaak. Yahi sabse common galti hai exam mein!
DO-178C ka signature concept hai MC/DC coverage — deadliest level (DAL A) ke liye. Iska matlab sirf ye nahi ki har condition true aur false ho, balki har condition independently output ko flip kare (baaki conditions fix rakh ke). Isliye A && B ke liye 2 nahi, n+1 = 3 tests chahiye. IEC 61508 mein quantitative target hota hai: PFDavg=λDUT1/2. Ye "divide by 2" isliye aata hai kyunki proof-test ke turant baad system theek hota hai aur interval ke end pe full risk — average beech mein. ISO 26262 mein ASIL nikalta hai S (Severity) × E (Exposure) × C (Controllability) se, aur ek smart trick hai ASIL decomposition — ek ASIL D requirement ko do independent modules mein baant do taaki dono kam rigor pe ban jaayein. Bottom line: ye standards bug-free guarantee nahi dete, balki failure ka probability acceptable level tak laate hain.
