Pushing the Frontier of Risk Assessment and CAPA

New technologies require new tools

Published: Thursday, September 17, 2015 - 00:00

Technology is evolving more rapidly than at any point in history. We tend to think of seminal moments in history, such as the inventions of the steam engine or powered flight, as literal moments in time.

The reality is that the invention of the steamboat by Robert Fulton was only possible because of a long sequence of successive inventions on which he was able to capitalize to change the world. If the ability to cast and temper iron had not been improved over the ages, Fulton could not have repurposed the concept used to pump water and compress air into the piston, cylinder, and crankshaft. Fulton’s brilliance was harnessing known physical laws to use steam to develop and transfer energy into work.

The steam engine appears with a specific date for its invention on those illustrated timelines so fondly recalled in high school history books. Depending on your perspective, it could very well look more like a Gantt chart, starting with man’s development of fire.

News media outlets are filled with stories about technological breakthroughs, including self-driving cars and drones being tested to deliver everything from pizza to life-saving medicine. After gushing about a new development, they will follow with a disturbing sequence focusing on the potential harm and disruption that might be caused with the invention by persons with ill intent.

This juxtaposition of oppositional ideas causes weird images to flit through my mental field of view. I can see Buckminster Fuller excitedly high-fiving Alexander Graham Bell for their new ideas, while a legion of historical villains are sneaking up on them ready to mount handguns on drones and hack self-driving cars onto railroad tracks in the path of high-speed commuter trains.

Timelines for world-changing inventions are littered with dead and injured early adopters of new and untested technologies: Steam engine boilers exploded regularly, and one of the Wright brothers sustained injuries that physically compromised him for the rest of his life. I doubt that the public would tolerate similar missteps testing drones and self-driving cars. Even though we seldom blink at the number of auto fatalities each year, I suspect that the first self-driven car fatality, especially if it involves an innocent third party, will almost certainly trip a deployment and adoption circuit breaker of some sort. The technology will eventually take root and thrive in spite of the deaths and injuries but a crash of a commercial airliner attributed to a drone will be a very tough consequence for the public to get past.

I don’t think we are less risk-aversive than our forbearers, but I think we take comfort in not connecting the dots between the emerging technology and unintended deaths and injuries to early users. Almost all of us drive cars with keyless ignitions without concern about computer-savvy car thieves. Data are just starting to emerge related to deaths and injuries from cars engines running in closed garages because drivers forgot to turn off their vehicles. This will almost certainly be addressed in a uniform fashion as a lesson learned for the next car design, but no one will stop driving keyless ignition vehicles based solely on this data.

The hidden burden falls on development and engineering groups and their quality professional counterparts to improve their existing toolset for identifying and mitigating risk.

There’s an enormous amount of pressure to develop and be first to market exciting technologies while assuring reliability and protecting users and the public. Those are pretty much oppositional imperatives and difficult to balance. Methodologies (e.g., failure mode and effect analysis) and tools (e.g., fault-tree analysis) to determine root cause analysis and risk-based corrective actions are penetrating beyond the typical industries we think of (e.g., automotive, aerospace) when discussing conventional risk management tools.

My professional experiences tell me that we’re overdue for a breakthrough that will dramatically improve the current risk assessment and mitigation toolkit. Improvements in virtual modeling and increased capacity in advanced data mining will almost certainly result in faster, more accurate, and more dependable preventive tools. There is absolutely no way that technology or the public demand for the next cool product will slow down. Those in the business of assuring that new technologies are safe and dependable, within the boundaries of acceptable risk, need to up their game and rise to the challenge.

The next generation of design and quality professionals are cutting their teeth with the old toolkit and mentally identifying shortcomings to determine brave new ways of addressing them.

All of us should be prepared to embrace the tools necessary to facilitate the pace and evolution of new technologies.

About The Author

Mary McAtee

Mary McAtee has been a member of the Siemens organization for more than 20 years. She is a 40-year quality professional specializing in reliability engineering for semiconductor and nuclear devices. McAtee is an exam-qualified lead assessor for ISO 9001, ISO 14001, ISO 13485, IATF 16949, and TickIT. She has lead several organizations to successful registrations to various standards and has written and presented on the topic of compliance and quality extensively over the years. She is working with organizations in the United States and Europe to develop a broader uniform interpretation of primary norms and compliance standards.