High Velocity Innovation at VDL ETG

Making crucial design decisions early in the concept phase of a project, may lead to severe cost and time overruns when these decisions are based on assumptions and/or incomplete knowledge. This causes redesign, re-development, and re-testing plus frustration across development and management.

In February 2023, VDL ETG started a pilot project with Rapid Learning Cycles (RLC), a framework for knowledge-intensive projects in the early phases of development. VDL ETG wanted to get more control and visibility of these types of projects that include innovative design concepts and therefore a number of risks and unknowns.

VDL-ETG’s requirements

VDL-ETG defined a set of requirements for the pilot project. The main objective was to get better grip on progress in the pre-concept phases of knowledge intensive projects. The pilot was planned for a period of 4 months. The project manager Frank Romberg and system architect Gerrit Oosterhuis participated in the pilot, joined by 10 lead designers. An application called Targetprocess was used to manage the project and provide visibility and transparency.

The RLC Pilot Project

Rapid Learning Cycles (RLC) is an agile product development framework tailored for the development of physical products. It is an intuitive framework that can easily be adopted by a development team.

The pilot started with a 2-day training course in which we directly applied the RLC theory to kick-start  the pilot project within VDL ETG:

1. Learning the key elements of the RLC framework.
2. Exercise with the key RLC elements on the project at hand.
3. Using Targetprocess to capture the key RLC elements.

These key artifacts of RLC are:

1. The Core Hypothesis
2. Key Decisions
3. Knowledge Gaps
4. Learning Cycle Plan

Core Hypothesis

During the RLC training the development leads collaboratively created the Core Hypothesis. By agreeing with the team what the customer need is, which new technology concept will satisfy this need and what the business value is, everybody shares the same understanding of what the aim is for the project.

Key Decisions

Based on the Core Hypothesis the team started to brainstorm on possible Key Decisions. A Key Decision is a decision or choice that has a high impact on the success of the product, but the team doesn’t have the knowledge to make that Key Decision with confidence.

At VDL ETG we made a distinction between decisions in the concept phase on System level and Module level. Decisions at system level were by definition Key Decisions. Module level decisions which cannot be made without affecting the concept, were also regarded as Key Decisions. All other decisions and choices were regarded as regular, non-key decisions.

Knowledge Gaps

Knowledge Gaps are things that the development team needs to know to make a Key Decision with confidence. Typically, a Key Decision has 1 or more Knowledge Gaps.

Knowledge Gaps are closed in 1 or more iterations of 3 weeks. After each iteration, finished Knowledge Gaps are presented and shared during a Learning Cycle event, a meeting with the team to share knowledge.

Typically, a team will have more Knowledge Gaps than they can close. Therefore, Knowledge Gaps are prioritized. For those Knowledge Gaps that cannot be closed, a Risk description is created. The severity of each Risk is assessed based on Impact and Chance.

At VDL ETG, it became clear that we did not only have a lot of Knowledge Gaps, but also a lot of Key Decisions to be made.

During the pilot project, we made some adjustments, within the boundaries of the RLC framework, to accommodate the dynamics in the project.

Learning Cycle plan

All identified Key Decisions and Knowledge Gaps were planned into the Learning Cycle plan. This plan makes it extremely visible when Key Decisions and Knowledge Gaps are planned and who is responsible for the work.

The Learning Cycle plan, Key Decisions, Knowledge Gaps and Risks were managed by Targetprocess.

The information in Targetprocess provided real-time insight into the planning of the work. It also allowed for the plan to change. As a result, the team was able to transparently show which work needs to be done, by whom, when it is planned and to what decision it is connected.

This transparency made it easy to create a common understanding between the customer and VDL ETG about the challenges and progress in this phase of the project.

Conclusions and observations

“The Rapid Learning Cycles framework is very intuitive and doesn’t require extensive training. This allowed the development team to focus the majority of its time on the content of Key Decisions and related Knowledge Gaps”, said Frank Romberg. “The Learning Cycle plan provided insight into who does what and when, but also if deadlines were feasible and realistic”.

Formulating Key Decisions and Knowledge Gaps

During the training the development team brainstormed on Key Decisions (KDs) and Knowledge Gaps (KGs) and quickly wrote down one-liners for all of them. These were explained to the group to ensure a common understanding.

Although at first it seemed that the formulation of KDs and KGs was sufficiently clear, later on there was discussion about the exact meaning of some of the KDs and KGs.

We learned that it is important to spend extra time with the development team on formulating KDs and KGs to avoid any confusions later on in the project.

Explicit decision making

Making Key Decisions with confidence is what Rapid Learning Cycles is all about. That also means that Key Decisions have to be made explicitly and are unambiguously formulated. Writing the Key Decision collaboratively with the team may seem overkill but does provide extra value: Everybody has the same understanding and agrees with the Key Decision.

Iteration length

The project had tremendous pressure to deliver on time and live up to customer expectations. As a result, we decided to have an iteration length of 2 weeks.

The module teams worked extremely hard and delivered a ton work every iteration. This created a challenge for the leads of each module team to be able to review Knowledge Gaps and Key Decisions before the meetings.

We solved this by changing from a 2-week to a 3-week iteration length. The last week could only be spent on peer reviews, approvals and preparation for the Learning Cycle and Integration Event.

When to apply Rapid Learning Cycles?

RLC focuses on pulling knowledge forward and pushing decisions later.

Mapping this to VDL ETG’s Product Generation Process (PGP), makes RLC specifically suitable in the Initiation, Definition and Concept phases. It makes this fuzzy process at the start of a product development project extremely transparent and visible. Progress is always clear and easy to communicate with Targetprocess.

Rapid Learning Cycles and Systems Engineering (SE)

VDL ETG is applying Systems Engineering from the Design phase of the Product Generation Process onwards. The question arises, how do RLC and SE connect?

Although this requires more investigation, it looks like the natural connection between RLC and SE is on the decisions. In RLC we have Key Decisions, and these can be translated to Design Decisions in SE. Design Decisions in SE will have associated requirements to specify the solution in line with the design decision.

When not to apply Rapid Learning Cycles

Rapid Learning Cycles is especially useful in the early stages of product development when there is a lot of uncertainty. The other way around is also true. So, if all requirements are clear for a product and it is also clear how to develop the product, then a traditional project management approach are more appropriate than a Rapid Learning Cycles approach.

Practical information

If you are interested in learning more about Rapid Learning Cycles, if you want to identify if Rapid Learning Cycles might be applicable to your project, contact Lizet Bary (l.bary@synergio.nl) or Edwin Schumacher (e.schumacher@synergio.nl) from Synergio to learn more about Rapid Learning Cycles in general.