2015-02-23

Green Software Project: Final Report published

The preprint of the final report of the project: “Establishing and exploiting potentials for environmental protection in information and communication technology (Green IT), Subproject 3: Analysis of potentials for optimizing software development and deployment for resource conservation” gives a good overview of the latest state of the art of the research activities in the field of green software and its engineering. The English version of the report consists of the following three parts: 
  • Starting points for resource conservation
  • Methodological challenges and approaches
  • Recommendations for actions
Additionally to these, the German version starts with an analysis of current trends in ICT. The following short summary of the report is based on the full German version.



Analysis of Trends

In order to set a basis for finding starting points for resource conservation in the area of ICT (following part), Hilty et al. identify emerging trends in the way of using ICT products and in software architectures. They point out that the trend of mobile Internet access on the one hand comes along with an optimization in energy consumption since the battery of mobile devices is limited. On the other hand the trends lead to an intensive usage and shortened lifespan of mobile devices. The  trend of “always-on” supports other tendencies like Apps (efficient programming but huge data transfer and storage) and Software as a Service. All of these trends do support a resource consuming usage of ICT products. Above, spam causes a high energy consumption and objects to many hardware capacities. The “traditional” Internet user especially consumes energy, needs resources and causes data traffic in using social networks and high resolution media. Here, online games, IPTV, 3D graphics and videos come to mind. Regarding Cloud Computing, the authors of the report see the potential of better hardware utilization but make the reader aware of rebound effects. 
Hilty et al. explain all of these trends in view of the resource consumption and explicitly point out the impact of it.


Starting Points for Resource Conservation

Based on the analysis of trends the report describes different starting points for resource conservation in the context of application software, and data centres as well as some additional points to that.

Regarding application software the following advices are presented:
  • Selectable image resolution (minimum resolution by default)
  • Mobile web preferably via WLAN (as the technology that is not that resource intensive)
  • Applying the “app principle” more broadly (apps are reduced on their main functionalities and thus need less hardware resources, this ideas could be transferred to stationary devices)
  • Implementing web-based applications efficiently (SaaS, Cloud)
  • Demand-adaptive software (modular, context sensitive) 
  • The role of open source software (pros & cons)

In the context of data centre Hilty et al. see additional points:
  • Dynamic predictive load management (optimal capacity, monitoring)
  • Information and data management (less data rate, energy efficient storage media)
  • Data compression and data deduplication
  • The challenge of a heterogeneous data center market

Additional to the aspects belonging to software applications and data centres, attention should be paid to:
  • User behavior in social networks (shorter usage time)
  • The polluter pays principle
  • Free up existing knowledge

Methodological Challenges and Approaches

In the final report the authors distinguish those aspects that still need to be addressed on the one side and existing approaches on the other side. Caused by the multifunctionality of software products the definition of functional units is still missing. This point will be one of the future challenges. Additionally, the fact that software is always a part of a complex ICT system with many influencing variables and that usage scenarios are still missing measurements of the energy consumption of software need to be further developed. Indeed, there already exist some approaches in this context when it comes to measuring single aspects of the ICT system.
Although software products are not yet established in the context of environment-related standards, there are some studies regarding CO2 emission of software (GHG, GeSi). However, a standardized definition of non-functional requirements of green and sustainable software products and of the whole software development process as well as guidance notes, checklists, and specific recommendations are still missing.


Recommendations for Actions 

At the end of the report, Hilty et al. list recommendations for actions that can be dedicated to four directions: research and standardisation, consumer-oriented, software developer-oriented and training and professional development. The first part addresses the lack of methods, standards, and requirements. All of them need to be defined to be able to regularly run standardized software tests in order to analyze the energy consumption and find out about the reasons for the consumption. The consumer should be informed, e.g. by developing a label like the Blue Angel for software products. Additionally, the environmental friendly efforts of users could be supported by recommendations for configurations of software that consume less energy. Software developers should also be supported by checklists, best-practices and guidance notes how to develop green software in a green way. Finally, the aspects of green and sustainable software engineering needs to be integrated into studies and advanced training. Here, appropriate teaching material should be developed.

Conclusion

Overall, the report gives a good overview of existing approaches and challenges in the context of green software (engineering). The authors touch on the different aspects and reference many literature. In order to get an additional impression of publications, one should read literature reviews1,2,3,4 as a supplement to this report.
Naming and describing existing projects and stakeholders in the addressed field would be an interesting completion to this report. Here, one has to draw on corresponding conferences, program committees, calls etc.

Complying with the title of the report “green software”, the authors only dwell on the ecological pillar of sustainability.


  1. Ahmad, R., Baharom, F., Hussain, A.: A Systematic Literature Review on Sustainability Studies in Software Engineering. In:: Proceedings of KMICe 2014
  2. Hilty, L.M., Lohmann, W.: An annotated bibliography of conceptual frameworks in ICT for sustainability. In: Hilty, L.M., Aebischer, B., Andersson, G., Lohmann, W. (eds.): ICT4S ICT for Sustainability. Proceedings of the First International Conference on Information and Communication Technologies for Sustainability, ETH Zurich, February 14-16, 2013, pp. 288–300. ETH Zurich, University of Zurich and Empa, Swiss Federal Laboratories for Materials Science and Technology, Z├╝rich (2013)
  3. Penzenstadler, B., Bauer, V., Calero, C., Franch, X.: Sustainability in software engineering: A systematic literature review (2012)
  4. Penzenstadler, B., Raturi, A., Richardson, D., Calero, C., Femmer, H., Franch, X.: Systematic mapping study on software engineering for sustainability (se4s)—protocol and results. Institute for Software Research, University of California, Tech. Rep. UCI-ISR-14-1 (2014)