Another milestone in life, my teenage daughter is learning to drive. I’ll freely admit to being a bit anxious about it. I recall her and her younger siblings playing with toy cars, and I don’t remember it ending well for the cars…
My training as an engineer has taught me to seek out potential hazards and look for ways to address them. I have every confidence in the ability of my daughter to safely navigate the roadways of Houston. It’s everyone and everything else that worries me. The focus on driver safety, much like the focus on safe and sustainable industrial operations, is to apply appropriate safeguards to hazards in order to minimize the possibility of catastrophic consequences. Crumple zones, seat belts and air bags are akin to pressure relief valves, interlocks, and safety instrumented systems. These safeguards are intended to be independent so that there is no single point of failure. In industrial facilities, these safeguards are often taken as an IPL, or Independent Protection Layer.
IPLs are implemented to reduce the risk of a hazard becoming a major event. In industry, each IPL must be designed and maintained to function with a very low probability of failure. The challenge within industry is ensuring that the design assumptions match the reality of operations, and that the IPL systems are properly tested. Immense amounts of effort go into the design and maintenance of these systems.
What if I were to apply the same rigor to the vehicle my daughter drives? What’s the probability of failure on demand of a seat belt? Does she wear the seat belt? Are the airbags part of the nationwide recall? What’s the expected frequency of activation of the ABS system, vs. the actual demand rate applied by my daughter? Most importantly, how can I verify? I trust, but still want to verify.
I can inspect the fibers of the seat belt for wear and tear on a monthly basis. I can check the tire pressure, break wear, etc. I can tap into the electronic “brain” of the vehicle and collect information on my daughters driving habits (not unlike some insurance carriers). Just like my experience within industry, I can spend countless hours generating volumes of information through manual or semi-manual means. It can be a herculean task to simply collect all the relevant documentation and verification information required by IEC 61511. A significant cost savings can often be had by streamlining the compliance process while maintaining existing plant processes and standards.
Once the data is collected, the challenge is to make it truly useful. Do I need to have a conversation with my daughter because the wear rate on the breaks is statistically deviant from the expected wear under typical driving conditions? How about a conversation where the seat belt wasn’t engaged for ten minutes after starting the car, while the vehicle traveled two miles from the house? These are great conversations to have, but are only effective if they are timely. Industry has these same challenges, the time it takes to collect all the relevant data from the myriad data sources and put all the pieces together often prevents timely review of the information to make it truly useful.
When my daughter asks to borrow the car, will I have the right information to make the best possible decision? What information would I need? Do I need brake pad thickness trend data to detect statistically out of control wear, or will a simple “they don’t squeak” suffice? In industry, the trend is to utilize auditable historical information as opposed to a simple “pass/fail” test. This dramatically increases the complexity and effort required to gather the data in order to convert it to usable information. It might not be enough to say “the valve closed.” Instead, I’ll have to prove that the valve closed in the required amount of time within the limits of statistical control.
At the end of the day, I definitely want to ensure the safety of the vehicle my daughter drives, just like we want to ensure the safety of the processes we operate. The challenge is using efficient, reliable methods to collect the data and turn it into actionable information.
When it comes to plant safety and understanding the availability and performance of your IPLs, do you automatically have all the information you need when you need it?