The Hybrid-Agile Design of Experiments Methodology

K Gradeci,M Sletnes

doi:10.1088/1742-6596/2069/1/012039

Abstract

A DOE (Design of Experiments) is the laying out of a detailed experimental plan in advance of doing the experiment. Optimal DOEs maximize the amount of information that can be obtained for a given amount of experimental effort. The traditional DOE methodology is waterfall-type methodology implying a sequential and linear life-cycle process. The success of the experiment and usefulness of the results are highly dependent on the initial experimental setup and assumptions, and does not allow to go back and change something that was not well-documented or thought upon in the design stage. The fast-changing software development industry have made it understandable that the traditional waterfall methodology for developing systems, which follows similar patters to the traditional DOE, lacks the agility required for developing robust systems. These limitations have triggered the development of agile: a type of incremental model of software development based on principles that focuses more on flexible responses to change, instead of in-depth planning at the design stage. This paper proposes the hybrid-agile DOE methodology – a methodology that incorporates agile principles in traditional waterfall DOE methodologies – to design effective experimental layouts that allow for improvement during the experimental trial process. The methodology is applied to the natural ageing of adhesives tapes for building applications. This methodology can overcome traditional DOE, by adding agility in the whole process, especially in cases where the investigated products lack prior information and are characterised by large variability.

Highlights

This work was conducted as a part of an R&D project aiming to develop robust test, evaluation and prediction methodology for ensuring durable adhesive airtight solutions for energy efficient buildings
This paper proposes the hybrid-agile Design of Experiments (DOE) methodology – a methodology that incorporates agile principles in traditional waterfall DOE methodologies – to design effective experimental layouts that allow for improvement during the experimental trial process
The traditional Design of Experiments (DOE) methodology is based on waterfall principles implying a sequential and linear life-cycle process where the setup is selected and fixed before the experimental trials starts

Summary

Context

There are several aspects of adhesive tape durability testing which makes it an interesting and challenging case for efficient design of experiments (DOE): a) The inherent complexity of the airtightness solution: Each air tightening joint consists of a substrate (typically membranes and/or structural building components), and a tape which consists of a (often multicomponent) carrier/backing and a pressure sensitive adhesive Together they make a complex composite, where each constituent layer, and the interaction between the layers have a different response and sensitivity to climate exposure stress factors.; b) The high level of co-dependence of the factors in the experiment – e.g. the peel strength of a tape cannot be measured independently of a substrate.; c) The unsuitability of known DOE methodologies such as fractional factorial design due to the categorical or stochastic nature of the input factors.; d) The lack of detailed knowledge on the composition of the different adhesive tape products on the market due to trade secrets.; e) The possibility of unpredicted tape/substrate interactions leading to unusual degradation [4].; and f) The diverse and stochastic climate conditions expected to influence the adhesive tape durability in actual applications. Showcase the proposed framework with an example of a long-term natural ageing of adhesive tapes for building applications

Waterfall methodology

Agile methodology

The proposed hybrid-agile design of experiments methodology

Conclusions