Product Operations Program ManagerBarcelona, Cataluña, Spain
Or, on the contrary, an Agile approach fits better? Share your experience and/or thoughts. Saving Changes...
Sort By:
Wade HarshmanScrum Master| GDITIndianapolis, In, United States
I've heard good Agile coaches state that Agile only has an advantage over traditional approaches (waterfall) when the project is highly complex or contains a large degree of unknowns. In cases where the project is well-defined, waterfall is perfectly adequate and proven approach.
Those conciliatory statements stood out to me in contrast to the typical Waterfall vs Agile wars that we often see.
I've had the opportunity to work on some R&D projects. In my case, the projects followed a waterfall approach: we need to discover A and prove B before we can test C, etc. This also fit the structure imposed on us by regulatory authorities. I'd love to hear from someone who has had a similar project in an Agile organization. Saving Changes...
It used to be my favorite project management question. Even did a little research but couldn’t find any practical solution. It applies to any innovation project. Or even to the innovation tasks on a not very innovative project (QA and testing activities). It comes down to the amount of unknowns in the project. In your R&D example, what we usually don’t know are:
- How long will it take to find a solution?
- How long will it take to test? How many times do we have to repeat a process?
- How many times we have to test a potential solution to find out it didn’t work and we have to restart?
- When are we going to stop trying if a competitor finds a solution earlier than us?
- And the list goes on
My take on is if we know with a good certainty what the end result has to be and how we are getting there, the better we can estimate and a waterfall approach would still work. I used to do a triangle estimate with worst case, best case and most likely measures.
I really like to hear from the people who actually had to deal with this in the industry. Saving Changes...
Great question. I work in the medical device field -- a counterpart to the pharmaceutical area -- and my experience with R&D development has been that it mostly follows the waterfall approach. As Wade points out, the development path follows a series of logical steps. However, the research path has greater flexibility which may fit more with the agile approach.
Once a product is moving through the development phases towards a commercial product, it needs to meet certain requirements for regulatory purposes. Generally, the developers must provide evidence that established processes were followed. This evidence customarily is documentation. And, there are many efforts to standardize documentation across the industry in all phases of product development.
The concept of the waterfall is even embedded in FDA guidance to manufacturers for Medical Devices (http://www.fda.gov/downloads/MedicalDevices/.../ucm070642.pdf) which talks about the application of design controls to the Waterfall Design Process. This essentially establishes the documentation that is required at the end of each design phase.
I'm no expert on Agile, but I would not discount the possibility of using the Agile approach in the research / early feasibility phase. This may be a way to facilitate quick decision making when there are many possible development paths as often this is the case in science.
This has been my experience, but I am eager to hear from others on the possibility of using Agile in science. Saving Changes...
Product Operations Program ManagerBarcelona, Cataluña, Spain
Thanks everyone for the very insightful comments. I believe that we look forward to gathering more opinions/thoughts/comments, so keep'em rollin'! Saving Changes...
It depends - as always. As Romiya points out, in a regulated environment, even R (of R&D) follows a waterfall approach. In other environments, I see more and more agile approaches (in former times "agile by accident", now more "agile on purpose"). Recently I heard, that even for medical devices some agile approaches started - within FDA regulations (within a phase). As Wade noticed: The more complex or unknown the project area is, the more agile is a good approach - early feedback, early reaction possible. This saves time and money on the way to solution. Research works that way since a long time - even if they didn't call it "agile" - more often "trial and error" :-). Saving Changes...
Karthik RamamurthyAuthor, Say YES to Project Success| Founder KeyResultzChennai, Tamilnadu, Tamilnadu, India
Eduard: As Rolf Dieter Zschau points out beautifully, the answer depends on the environment.
I would add another point of dependency: Knowledge and experience with Agile within the performing organization and the client.
Criticality of the project is another important factor.
After all, you would, any day, certainly stick with "tried and tested" approaches on a critical project rather than take a risk with agile with a team that is not very mature with the methodology! Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
I have used both in more than concrete situation. First, as you pointed out in your statement, waterfall and agile are approach totally independent of the method and life cycle you will use. In my personal experience, the critical point to decide about what approach to use was the type of product to create and how well product characteristics and features were known in advance. That has determined the pace and rythmin of the R&D process. Saving Changes...
Pravin KajarekarPM III| IBM India Private LimitedMumbai, Maharashtra, India
An Hybrid approach is better. We define long term goals and use Agile to achieve each of the milestones. The availability of funding will probably defines the best approach. Saving Changes...
Community Champion
