RHEA’s experts have been using concurrent design for nearly two decades to accelerate the early phases of complex engineering projects such as space programmes, defence systems, factory design and luxury yachts.
In this new blog series, they share their insights into how you can ensure concurrent design sessions are rigorous and the outcomes are high quality.
By Gwendolyn Kolfschoten, Concurrent Design Expert
3. Validation and doing the maths
Validation is a standard practice in system engineering, but it is still mostly a planned and separate activity at the end of a task when something is finished. For concurrent design, we need a more continuous and iterative means of validation, and preferably one that is quantitative and driven by data.
To make a quantitative validation, we need data; however, that is often not yet available at this stage and therefore we use estimates.
Engineers are used to calculating the properties of a system, but coming up with an early estimate, before rigorous calculations are possible, is more difficult.
Rough estimates have value
Doing early validation based on rough estimates can feel uncomfortable, especially for engineers, whose job it is to make sure calculations are correct. Nevertheless, our experience shows that early estimates, however inaccurate they are, reveal many insights and create considerable understanding about the key trade-offs and the relative constraining nature of different design parameters.
For instance, in one project a key constraint for a newly designed ship was the size of the harbour. At first, the impact of this limitation seemed minimal. However, we translated this into the maximum length and width of the ship to determine the likely centre of gravity, and consequently a maximum mass – this turned out to be much lower than the initially targeted mass, and therefore became a key design driver.
Such initial checks come with much uncertainty: in this case the shape of the ship was not yet known, so we could not yet model or calculate an exact centre of gravity, the target mass was not yet based on realistic values, and so the comparison was very crude. However, if the comparison showed the difference to be more than 50 tons, it was safe to assume that this was a critical issue. Even if it turned out to be a difference of 70 or 30 tons, it was a big problem, and would require choices to be made in the design.
Helping the team to estimate
Making these first rough validation checks often causes much resistance from the team: the information is not available or incomplete, based on assumptions and probably not correct. However, when we get a feeling for the order of magnitude, it helps the team to see what constraints they need to prioritize.
Here are some ways to motivate the team to make these first estimates:
- Explain why. Use the story above, or other examples, to make the team understand why it is important to make estimates early, and not to postpone them until they can be made more accurately.
- Be specific in the assignments of what experts need to calculate. It is not sufficient to ask the team to provide input for a mass balance, or for a power budget – we need to be more specific about the way we approach such a high level calculation, and how experts should make the estimate. (You can read more about the value of estimates and margins of error in a previous blog post ‘Be Driven by Data’.)
- Set a solid deadline for the calculations; because they can always be improved, they can always be postponed unless you give a deadline.
- Indicate that you need a figure ‘to the best of your knowledge now’ in order to make the experts comfortable with the idea that it is OK that these are estimates and therefore not perfect answers.
- Fill in the gaps. When experts come back ‘blank’, go through the process and ask them where they got stuck. You can also ask others to help think about an approach to estimate the missing values.
Find out more
In this series of blog posts on concurrent design, we present techniques to improve rigour and quality. These offer a toolbox for concurrent design team leads to support their teams effectively in improving and testing the design. We cover aspects such as how to handle validation when working with rough estimates, how to zoom in and focus without losing a holistic view, and how to keep the team on track and contributing positively.
Follow RHEA Group on LinkedIn to be alerted to new posts in this series.
Catch up with our previous blog series on 10 Success Factors for Collaborative Design.
