FairGA: Fair Genetic Algorithm - Beyond Resource-oriented Sustainability for ICT Products and Services

TL;DR

FairGA is a novel genetic algorithm designed to promote nondisposability in ICT products and services, aiming to reduce environmental footprint by influencing unobservable processes during optimization.

Contribution

It introduces the FairGA, a modified genetic algorithm that incorporates nondisposability principles to manage unobservable processes and minimize resource consumption.

Findings

FairGA reduces resource extraction compared to traditional GA.

FairGA achieves comparable optimization performance.

Promising results in lowering environmental impact.

Abstract

The complexity of ICT products and services has brought them to the level of disposable `species'. The combination of the race to optimal performance and disposability has resulted in considerable footprint and impact. Although approaches such as increasing efficiency, reducing the total cost of ownership, life cycle assessment and management, and circular economy have been put forward to manage and reduce the footprint and impact, the complexity of processes involved and especially invisibility of key but unobservable processes has resulted in some lower bounds for minimal achievable footprint. In this work, a modified approach to the Genetic Algorithm is proposed in order to introduce the notion of 'nondisposability' to the ICT products and services in order to implicitly influence and manage unobservable processes, and ultimately reduce the overall footprint and resource consumption. The proposed genetic algorithm is called FairGA, and it is compared with the traditional Genetic Algorithm against standard optimization functions. Also, the impact of the FairGA on the resource extraction has been illustrated with promising results.