söndag 12 mars 2017

Evaluating the (un)sustainability of Computing Projects


Anton Lundström and I submitted a paper to the upcoming Computing within Limits workshop (conference), "Evaluating the (un)sustainability of Computing Projects: A Taxonomy of Value for Sustainable Computing". The paper builds on Anton's master's thesis that I was both the principal and advisor of, "Developing a Taxonomy for Sustainable ICT: An exploratory study of the feasibility of a taxonomy for sustainable ICT" and that he wrote last year.

The basic question we discuss is, "how do you know if a computing project is or isn't sustainable"? According to what principles can you evaluate the sustainability of a computing system? The answer to that question is a proposed framework, or a taxonomy for discerning what is and what isn't sustainable computing. Both the thesis and the paper in it's turn build on a paper that was presented at the first (2015) Computing within Limits workshop by Kentaro Toyama, "Preliminary Thoughts on a Taxonomy of Value for Sustainable Computing" (pdf). We have developed Toyama's "preliminary taxonomy" and hope the result of our efforts will be useful for others. Here's the abstract:

Evaluating the (un)sustainability of Computing Projects: A Taxonomy of Value for Sustainable Computing

Anton Lundström & Daniel Pargman
KTH Royal Institute of Technology School of Computer Science and Communication Stockholm, Sweden

Radical measures are needed to reach the 2 degree climate goal and achieve a sustainable society. Computing is often invoked as a means for achieving these goals and it has been estimated that the abatement potential of computing between now and 2030 is around 20% or 12 Gton of decreased CO2 emissions [8]. What is often forgotten is however that the infrastructure for computing also has negative impacts in terms of sustainability. Toyama [15] has proposed a “preliminary taxonomy” for classifying computing projects as a way of separating sustainable computing efforts from unsustainable ones. In this paper we explore the feasibility of Toyama’s taxonomy. We begin by describing how we developed the taxonomy to make it more practically useful and then conducted a study where we used the revised taxonomy. We evaluated four computing projects by interviewing project leaders and we use these interviews as a foundation for further discussions about the taxonomy. This resulted in yet another, third and final version of the taxonomy. We conclude that the final version is more practically useful than Toyama’s preliminary taxonomy [15], but that there are still challenges that need to be addressed. We end the paper by suggesting where future efforts should be focused.

Inga kommentarer:

Skicka en kommentar