I myself have not worked in industry yet, but I asked some family members who do about your question. PERT and CPM are used more implicitly in practice than their textbook form that you have mentioned above. A lot of medical device project phases have sequential tasks with dependencies, so CPM tends to be more commonly applied. These tasks can be design verification, validation testing, and regulatory submissions, and CPM can be applied here since these activities have fixed durations and strict ordering. Critical path analysis becomes very useful here as well, since it helps identify schedule risks. 

PERT is more helpful in earlier development of the project cycle or projects with more uncertainty. Examples of this would be novel device development or in-human studies. Optimistic, pessimistic, and most likely duration estimates make sense here since there are unknowns like FDA review cycles, clinical enrollment, or design changes. However, when uncertainty decreases, companies move to CPM-style planning. 

In practice, companies don’t label what they’re doing as CPM or PERT, but the logic behind both methods is embedded into the project management software. Microsoft Project or Smartsheet handle task dependencies and schedule risk analysis via the basic principles behind PERT and CPM. 

I am curious to see how everyone else has seen this applied to real settings. Do medical device companies teach and reference PERT and CPM, or are these concepts silently embedded in their work? How do teams account for uncertainty when building schedules? How can AI be integrated into schedule building most effectively? 

I have not formally worked in industry either, but I do have some knowledge of PERT and CPM in the medical devices industry, gained through conversations and internships. I certainly agree that PERT and CPM are rarely mentioned directly by name, at least in my limited experience. However, I will say that their principles are integrated into the way projects are executed. CPM definitely kicks in during late-stage development/post-design phases as regulatory, verification, and validation have clear dependencies/deadlines. Thus, missing a verification test required for submission, which is most certainly on the critical path, may delay an entire market launch. Thus, PMs leverage dependency mapping and critical path analysis, even if they don't call it CPM. Instead, I have seen the use of Microsoft Project, Smartsheet, and even Jira, all of which reinforce CPM logic through task sequencing and milestone tracking. 

PERT-style thinking is more common in early development, as noted in the previous post, when there is unavoidable uncertainty. I have seen PMs use both optimistic and pessimistic estimates here, since things like prototype iterations, clinical enrollment timelines, or FDA feedback cycles do not have fixed durations. Thus, even though PERT calculations are not formally done, the scenario planning and contingency buffering reflect this shared mindset. I am curious to see how teams decide what level of buffer is enough when uncertainty is very high. Is expert judgement relied on more, or is it historical data, or something else?