F𝗼𝗿 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗼𝗳𝗳𝗶𝗰𝗲𝗿𝘀 𝗮𝗻𝗱 𝗦𝗦𝗘𝘀 𝗽𝗿𝗲𝗽𝗮𝗿𝗶𝗻𝗴 𝗳𝗼𝗿 𝟭 𝗝𝘂𝗻𝗲 𝟮𝟬𝟮𝟲

Most critical control frameworks I have seen assume humans will behave consistently and infallibly. The science says they won't.

Ron McLeod's 2017 paper titled "𝙃𝙪𝙢𝙖𝙣 𝙛𝙖𝙘𝙩𝙤𝙧𝙨 𝙞𝙣 𝙗𝙖𝙧𝙧𝙞𝙚𝙧 𝙢𝙖𝙣𝙖𝙜𝙚𝙢𝙚𝙣𝙩: 𝙃𝙖𝙧𝙙 𝙩𝙧𝙪𝙩𝙝𝙨 𝙖𝙣𝙙 𝙘𝙝𝙖𝙡𝙡𝙚𝙣𝙜𝙚𝙨" lays out four "hard truths" of human performance and shows how each one routinely defeats well-designed critical control strategies.

Shout out to Ben Hutchinson (PhD) for originally alerting me to this paper. 

The strongest insights from this excellent paper are in the carousel. Here's what stood out to me:

• Most "human controls" in our bowties are actually 𝙨𝙖𝙛𝙚𝙜𝙪𝙖𝙧𝙙𝙨, not 𝙘𝙧𝙞𝙩𝙞𝙘𝙖𝙡 𝙘𝙤𝙣𝙩𝙧𝙤𝙡𝙨 (barriers). They support the real critical controls as they aren't reliable enough on their own to be one.

• Competence and training don't override System 1 thinking (Daniel Kahneman). A trained rail lookout walked in front of an oncoming train he'd already acknowledged. He had no doubt. System 1 doesn't experience doubt.

• Operators under pressure to make work easier will defeat your critical controls without ever recognising they're doing it. 

• If your CCM framework relies on "𝙩𝙧𝙖𝙞𝙣𝙚𝙙, 𝙘𝙤𝙢𝙥𝙚𝙩𝙚𝙣𝙩 𝙤𝙥𝙚𝙧𝙖𝙩𝙤𝙧 𝙛𝙤𝙡𝙡𝙤𝙬𝙞𝙣𝙜 𝙥𝙧𝙤𝙘𝙚𝙙𝙪𝙧𝙚", you don't have a critical control. You have an unrealistic expectation.

For officers of coal mines and metal mines & quarries, this connects directly to s47A and s44A, respectively. Due diligence isn't about whether you have critical controls listed. It's about whether the controls actually do what you're relying on them to do, under real human performance conditions.

The link to the full paper here: Human factors in barrier management

Check out my thoughts here: Hard truths

Happy to discuss what good looks like in practice. 

Scott Graham

Founder and Managing Director, Mineplex 

0400 820 250 

sgraham@mineplex.com.au