Automation built on observing real work

In this case the problem was not a lack of tools, but unrecognised patterns shared by many procedures.

Daily checks and multi-step diagnostic procedures required many manual steps. Each activity had its own procedure. Existing automations covered individual processes, not the elements shared by many of them.

I based the automation on four principles:

• Observing the real work — analysing how the work is actually done: the order of steps, the decisions taken and the operations that repeat, not only the procedure descriptions.
• Patterns shared across procedures — extracting mechanisms that recur in different procedures, such as repeatedly collecting and reconstructing the same data by hand.
• Automating whole sequences — replacing entire sequences of steps rather than individual actions.
• Information for the right recipient — results prepared for each engineer’s area of responsibility, instead of collecting the same data by hand every time.

I observed the work of more than 30 operators. On the basis of the patterns I identified, I designed and deployed more than 50 automations that replace whole sequences of steps previously carried out by hand.

They included a solution that automated the preparation and reporting of operational checks and supported the analysis of historical configurations. One of the automations replaced a sequence of manual steps that many engineers carried out every day.

The effectiveness of the first solutions led other operational areas to put forward further processes for automation.

Value: recovered operational capacity — automation carries out repetitive sequences of steps, and engineers get the information they need without collecting the same data by hand.