Data representation methods for environmentally conscious product design

Devarajan Ramanujan, Purdue University

Abstract

The challenge of holistically integrating environmental sustainability considerations with design decision-making requires novel representations for design and sustainability-related data that allow designers to understand correlations among them. Challenges such as (1) lack of suitable data & information models, (2) methods that simultaneously consider environmental sustainability as well as design constraints, and (3) uncertainty models for characterizing subjectivity in environmental sustainability-based decision making, pose serious impediments towards this goal. A wide body of previous research has focused on developing computational methods for modeling environmental impacts and indicators for products and processes. However, a significant majority of these methods stop short of providing a decision-making framework based on the calculated metrics. Additionally, most of these methods are built from the stand point of an environmental sustainability expert. They stop short of creating data representations that contextualize this information to designers and other stakeholders in product lifecycle management. Such gaps create a significant cognitive barrier which can prevent decision makers from integrating environmental sustainability considerations within product design. Imminent regulations that limit the environmental impact of products and services will compound these knowledge gaps as decision makers will begin utilizing existing tools without a well defined understanding of assessment results. More importantly, only using assessment-focused tools prevents decision makers from developing insights about correlations between environmental impact indicators and design parameters. Without this understanding, designers cannot apply findings from environmental assessment to design processes. Addressing these gaps necessitates methods for representing environmental sustainability data in a manner that makes it congruent to the design process. To this end, this thesis explores (1) novel data representations, (2) decision-making methods, and (3) exploration support tools that facilitate integration of design and environmental sustainability-related parameters for eco-conscious product design. We start our discussion by looking at existing tools and methods for integrating environmental sustainability assessment within product design. Next, we discuss decision-making models for eco-conscious design and methods for evaluating uncertainties in this context. Following this, we make the case for information visualization-based tools for environmentally conscious exploration of design alternatives. A consistent theme within this thesis is translating learnings from research into educational and industrial practice. Realizing a more sustainable world requires training engineers and students in concepts of decision-making for environmental sustainability. Along these lines, we conclude the thesis by discussing a guided discovery-based instruction framework for teaching environmental sustainability in existing undergraduate mechanical engineering curricula.

Degree

Ph.D.

Advisors

Ramani, Purdue University.

Subject Area

Mechanical engineering|Sustainability

COinS