Guest Column | July 9, 2021

How To Enhance "Operational Learning" In Biopharma

By John Wilkes (AstraZeneca), Clifford Berry (Takeda), Amy D. Wilson, Ph.D. (Biogen), Ashley Waddey (Biogen), and Jim Morris (NSF Health Sciences)

We appreciate the positive feedback from readers following our two-part series, “Human Performance in Biopharma Operations.” The series mentioned the importance of operational learning. Since several readers expressed interest in this topic, we are expanding on operational learning in this article – providing examples to contextualize operational learning and offering guidance based on our experience. We welcome your feedback and invite you to complete the short survey following this article.

We define operational learning as the capacity of an organization to understand how people and systems perform within the context created by the organization. It is learning about work from the perspective of the people performing the work.1

According to conventional wisdom in biopharma, operational learning is derived from multiple existing practices, including inspections, audits, deviation investigations, and gemba walks. As an example, an audit may stop with a finding that indicates people are not following a procedure. But that alone does not help you learn and cannot drive improved performance. What is often missing is the necessary understanding of the “how” for current performance, which can be found in the actions, assessments, and stories of the people who do the work. What sets exceptional organizations apart from the others is the ability to effectively engage the people who do the work as participants in learning – the manufacturing technicians, quality control analysts, warehouse associates who create success and safety almost every day despite often flawed and imperfect systems.

There are many ways you can improve your operational learning. However, before diving into the practices, we must emphasize the importance of establishing the right environment. The effectiveness of every practice described in this article relies upon people feeling safe to be candid about the realities of their work. Many in biopharma refer to this as “speak-up culture,” and it can only be realized through leaders creating psychological safety. Amy Edmondson defines psychological safety as a belief that one will not be punished or humiliated for speaking up with ideas, questions, concerns, or mistakes.2 Essentially, it is a shared sense of candor. Without the necessary psychological safety, people will rightfully withhold the details of how their work is done to protect themselves from sanction or reprimand.

A first step in enhancing operational learning is enhancing psychological safety.  Here are some pointers from Amy Edmondson on how to achieve psychological safety in the workplace.2

  • Create shared expectations and meaning regarding failure, uncertainty, and interdependence to clarify the need for voice and by emphasizing purpose.
  • Invite participation by acknowledging gaps, asking good questions, modeling intense listening, and creating forums for input.
  • Respond productively by expressing appreciation, destigmatizing failure, and maintaining an orientation toward continuous learning.

Complete and effective operational learning is enabled by a variety of practices, including leaders going to the gemba (work observation), post-activity learning (includes post-job reviews and after-action reviews), and learning teams. 

Leaders Going To The Gemba (Work Observation)

Operational learning at gemba is about leaders going to where work occurs to learn about the difference between work as imagined and work as done. Work as imagined refers to the various assumptions that people who do not perform the work have about how the work should be done. Work as done refers to how a task is actually performed by the people who do the work. These two terms are often used in discussing contemporary approaches to safety management.3

Understanding how work as imagined differs from work as done for typical work, across the spectrum of successes and failures, is essential for an organization to improve a sociotechnical system’s reliability and resilience, ultimately protecting people and product from harm. Deming is often attributed with stating, “Eighty-five percent of the reasons for failure are deficiencies in the systems and process rather than the employee. The role of management is to change the process rather than badgering individuals to do better.” The effectiveness of operational learning by leaders at gemba can be maximized by reminding ourselves of Deming’s words and by building the following practices into how leaders show up at gemba.

  • Leaders must actually watch a work activity. This is very different than walking the value stream, management by walking around, or auditing workplace conditions for inspection readiness.  A reasonable duration of time to watch a work activity is between 45 and 90 minutes.
  • A powerful way to show up at gemba is with what is referred to in Zen Buddhism as shoshin, or a beginner’s mind. Remind yourself that “it’s hard to learn if you already know.”4
  • This is not a process conformation with the Tayloristic assumptions of force fitting “one right way.” Rather, watch work with an appreciation of how workplace adaptation and variability mostly create success and safety. In cases where the blue line (work as done) is far from the black line (work as imagined) and edging up too close to risk, take that as learning that the system must be adjusted by adding capacity and/or controls, not that people must be adjusted.
  • Instead of asking questions to test colleagues, ask questions that you are truly curious about and do not know the answer.
  • Refrain from telling or correcting. Instead, practice what Edgar Schein calls “humble inquiry”5 by asking “how” questions and do a better job of listening and acknowledging.
  • Human and organizational performance is about understanding how people interact with plant, processes, and each other as part of a system to manage risk and protect people and products from harm. The interactions of people as part of the system is the focus of your operational learning at gemba. A powerful systems-focused question to consider asking is, “Where in the task are you having to make do, improvise, interpret, or adapt to create success?”

Using the practices described above, a leader going to gemba at Takeda spent 60 minutes observing wave bag sampling and was able to learn about multiple opportunities where management of risk to product could be improved. One example was that two automation prompts were displayed by the human machine interface (HMI) on top of each other to execute different but similar equipment sequences. The manufacturing technicians explained that people had selected the incorrect prompt in the past and, in those situations, automation engineers were contacted to correct and restart the process. As a result of this valuable learning, the issue was added to an ongoing project to improve HMI automation in that part of the plant.

Post-Activity Learning (Post-Job Review & After-Action Reviews)

A post-job review is a method of self-assessment conducted after a work activity to solicit feedback from the participants. Essentially, the goal is to learn about how success or failure was achieved. Usually, this involves a short duration (5 to 10 minutes) face-to-face dialogue between the participants in an activity soon after the activity concludes. The participants discuss what went well and why, and also what can be improved and how.

Good post-job review practices include:

  • Looking for areas or parts of the task that were confusing, took longer than expected, or caused re-work, and
  • Assessing what contributed to task success (if it went better than usual), learning from the experience, and incorporating the learnings for future execution of the same tasks.

The after-action review (AAR) is a technique to make learning routine, similar to that used in sports where coaches and players meet shortly after a game to discuss the team’s performance. AARs are used by the United States military services and their use has extended to business as a knowledge management tool in order to build a culture of continuous improvement.6 The process may be formal or informal, may involve large or small groups, and may take minutes or hours.

Facilitate a dialogue by asking your team the questions below. Roughly 25% of the time should be devoted to the first two questions, 25% to the third question, and 50% to the fourth question.

Q1: What did we set out to do?

Q2: What actually happened?

Q3: How did it happen?

Q4: What are we going to do next time?

How can after-action reviews help you? At Biogen, we conducted an after-action review to learn more about a success. On one weekend we harvested three batches simultaneously, which had never been done before.  Since we did this successfully, we conducted after-action reviews with the teams involved to identify what enabled that success and to identify what we would do next time to ensure these practices would be sustained.  One action to come out of this learning was to put in place a “Harvest Playbook” that outlined all of the coordination activities that were required. This playbook continues to be owned by the teams and is expanded as additional items are identified to share for continue successful performance.

Learning Teams (A Fast & Effective Way To Learn More)

A learning team is an approach used to learn about the systems in which people work, from those who do the work. It is a facilitated, open discussion with those who often perform a task, were present for a specific event, and those who design and improve the work.1 Three goals of the process are:

  • Learn about the context of work
  • Bridge the gaps in communication and understanding between workers, designers, and managers
  • Improve the work process

This approach is especially useful when we experience a near-miss or an outcome we did not want or expect and want to make sure we understand how we got there so we can improve. Learning can be challenging because what might at first appear to be a simple story is usually not so simple; there are usually multiple conditions and factors to consider, and people may offer different and sometimes conflicting perspectives. Learning teams increase our understanding of how work is performed and the context of the systems people work within.  This increased understanding ensures that actions we take drive improvement more effectively and address what will be most impactful to performance.

How might learning teams help you? As an example, in 2020 at Biogen we identified a trend of over-pressurization events, generally leading to blown rupture disks. We already had an engineering team in place to review these trends and recognized that we needed to augment the learning by engaging those that performed the work. We started to conduct learning teams for over-pressurization events on activities that included manual manipulations. This brought a new dimension to how we understood the conditions leading up to these events, helping to identify more impactful changes (see Table 1).

Table 1

Before the Learning Team

Not much to fix…

After the Learning Team

Many ways to make this better…

A rupture disk burst on a chromatography system during a CIP cycle because a spool piece was not installed.


“Not sure there’s much to talk about, the operator just missed the step in the procedure.”

Design Improvement Opportunities Identified

  • Spool piece installation is manual and there is not a way to detect when it is not installed
  • Work requires back and forth between automation instructions and work instructions
  • Automation prompts are not specific to what is expected


Work Practice Improvement Opportunities Identified

  • Working on similar tasks on similar equipment around the same time
  • Expectations to work on other tasks during down time between steps
  • Practices regarding when spool piece is installed are not consistent between shifts


What is common about each of the above examples of operational learning is that each involves personnel performing work at the “sharp end” – where actions taken by the worker could have immediate and/or irreversible undesired consequences. By embedding these techniques into how we lead and manage work, it has been our experience that operations will realize a true step change in improvement and, importantly, also realize a significant change in the level of employee engagement and overall employee satisfaction.  

We hope this article has inspired you to take action to do even more – even better – than you already are.

Operational Learning Survey

Please take our brief survey here by August 31, 2021, to tell us more about what you are doing and what you would like to learn more about. We’ll compile the responses to share in a subsequent article on operational learning.


  1. Bob Edwards and Andrea Baker (2020). Bob’s Guide to Operational Learning: How to Think Like a Human and Organizational Performance (HOP) Coach
  2. Amy Edmondson (2019). The Fearless Organization: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth.
  3. Erik Hollnagel (2017). Hindsight25, Can we ever imagine how work is done?
  4. Amy Edmondson (2017). TED Salon: Brightline Initiative, How to turn a group of strangers into a team
  5. Edgar Schein (2013). Humble Inquiry: The Gentle Art of Asking Instead of Telling
  6. Darling, Parry, and Moore (2005). Harvard Business Review, Learning in the Thick of It

About The Authors:

AuthorJohn Wilkes is the human performance program lead at AstraZeneca. He has more than 25 years of experience in industry, with experience in manufacturing operations, operational excellence, quality systems, and quality control.


AuthorClifford Berry is director and head of business excellence for Takeda at its Massachusetts Biologics Operations site. He has been a human and organizational performance practitioner since 1999, with experiences in commercial nuclear electrical generation, electric transmission and distribution, and biopharma.


AuthorAmy D. Wilson, Ph.D., is director of global human performance at Biogen. She has more than 20 years of experience in biopharma manufacturing, with focuses on human and organizational performance, operational excellence, risk management, and technical training.


AuthorAshley Waddey is human performance lead, U.S. Operations, at Biogen.




AuthorJim Morris is executive director of pharmaceutical consulting services at NSF Health Sciences. He has over 25 years of pharmaceutical management experience in both plant operations and corporate offices in the U.S. and Europe. He has held positions as deputy director of QA/QC and regulatory affairs at Mass Biologics, director of QA/QC for the Biologics business unit of Cilag AG, and a number of quality assurance and manufacturing roles with Pfizer over a 16-year time frame, culminating as the head of quality assurance in Latina, Italy. His areas of recognized expertise include quality leadership development, supply chain auditing and managing audit programs, quality management systems parenteral product manufacture and compliance, and OTC product manufacture and compliance.