Tips for Mistake Proofing in IT

Poke-yoke or mistake-proofing is the idea of designing something so that it cannot be manufactured or used incorrectly. What can you do to implement mistake proofing in IT processes or your user interface?

  • Avoid using modes. If you do use them, make it utterly obvious what mode the item is in at all times.
  • Provide graphical, easy to understand status reports. Ensure that they notify users of the ongoing status and flag unique, novel or out of spec statuses.
  • When someone runs into a rule preventing an action like an access control limit, give clear guidance through the user interface on how to get permission, the proper process and/or how to contact an expert for guidance.
  • When someone repeatedly commits the same error, train but don’t shame.
  • If many users are making the same error, collectively train everyone because those making the error are not the only ones who don’t know how to do it right.
  • Design processes to be as simple as possible but no simpler. Simplify existing flows as much as possible.
  • When designing a new process for something different from a common process, use different starting steps so that people don’t accidentally follow the common process out of habit when.
  • Wherever possible, give people the ability to go back in the process if they accidentally start the wrong one.
  • The default solution for workflows should not be to add a new spaghetti line on the flow charge for every situation or exception.
  • Review your process flows to make sure they closely fit business processes instead of becoming a hindrance to them.
  • Demand as little data as possible from users, but make essential fields mandatory to progress.
  • Design unique controls and interfaces to minimize accidental selections of the wrong thing.
  • Have a clean, uncluttered user interface and reports so people can quickly find what they want.
  • Have clear indicators when something is wrong. Think of the example of statistical process control charts or green/red indicators on system health dashboards.
  • If an automated system is giving advice, have an obvious method for reporting the exceptions and problems.
  • Have subject matter experts with the authority and knowledge to handle the rare exceptions or ability to over-ride processes, as well as oversight so they don’t abuse it to improve productivity.
Filed under: An IE in IT

About the Author

Posted by

Tamara Wilhite is the IE in IT blogger for the IISE. She is a Six Sigma green belt with experience in IT, PDM software, the defense industry and recycling industries. She currently works as a freelance technical writer.