Sustainable Software Engineering 🌱
Edition of 2023/24
MSc in Computer Science, Delft University of Technology
Sustainable Software Engineering is an overarching discipline that addresses the long-term consequences of designing, building, and releasing a software project. By definition, sustainability covers five main perspectives: environmental, social, individual, economic, technical. This course focuses on the first three.
Software Engineering (SE) has long addressed sustainability by narrowing it down to economical and technical sustainability. However, our society is facing major sustainability challenges that can no longer be overlooked by software engineers and computer scientists. It was estimated that, by 2040, the ICT sector will contribute to 14% of the global carbon footprint. Hence, environmental, social, and individual ought to be part of the equation when it comes to design, build, and release software systems. The problem is far from simple, but we need expert computer scientists to bring sustainability into the core values of the next generation of tech-leading organisations.
This course covers a set of competencies needed to leverage sustainable software systems. It has a strong component on Green SE, covering techniques to measure and improve the energy-efficiency at any stage of the software lifecycle. Students learn state-of-the-art practices on energy efficiency and apply them in real software projects. Moreover, the course will cover core principles of empirical software engineering, and social and individual sustainability.
Organisation
Course code | CS4415 |
Brightspace đź”’ | https://brightspace.tudelft.nl/d2l/home/595278 |
Instructors | LuĂs Cruz, June Sallou |
Schedule | Mondays 10:45, Wednesdays 13:45, Thursdays 8:45. đź”— MyTimetable. |
Mattermost | Link shared on Brightspace. |
Zoom link | Link shared on Mattermost. |
ECTS | 5.0 |
Quarter | Q3 |
Format | Classes are optimised for in-person attendance; Online attendance is allowed in a case by case scenario. |
Examination type | Group Project 1 (40%); Group Project 2 (60%). |
Target audience | Students of the [M.Sc. in Computer Science]. |
Requirements | - Intermediate understanding of OOP languages; - Basic understanding of data analysis techniques. |
Learning Objectives
By the end of this course you will be able to:
- LO1. Measure software energy consumption.
- LO2. Automate carbon-awareness in software development.
- LO3. Discuss sustainability principles.
- LO4. Solve sustainability issues in real software projects.
- LO5. Propose innovative strategies to monitor software sustainability.
Outline
⚠️ Please note: Recordings are only available to the students because they are not edited and may include students’ interactions.
Class | Week | Summary |
---|---|---|
1 | 1 | Lecture. Course introduction. Sustainable Software: What, Why and How. đź“ŠSlides in-person |
2 | 1 | Lab. Measuring software energy consumption. Introduction to Project 1. 📊Slides 🎥Recording in-person |
3 | 1 | Lecture. Green Software Engineering — Part I: units of energy. 📊Slides 🎥Recording in-person |
4 | 2 | Project. Project 1 - steering meeting and formative assessment. in-person |
5 | 2 | Lecture. Green Software Engineering — Part II: Scientific guide for reliable energy measurements. 📊Slides in-person |
6 | 2 | Guest Lecture (Feb 22). “Following the Sun” to Mitigate the Carbon Footprint of Training AI Algorithms: is it Worth the Price?. 📊Slides in-person |
7 | 3 | Project. Project 1 - steering meeting. in-person |
8 | 3 | Lecture. Green AI. 📊Slides 🎥Recording in-person |
9 | 3 | Lecture. Green Software Engineering — Part III: Energy efficiency in mobile computing; carbon-aware data centres. 📊Slides 🎥Recording in-person |
 | 3 | Deadline for project 1 Friday, March 1. |
10 | 4 | Project. Project 2 - Description and kick-off. đź“ŠSlides in-person |
11 | 4 | Guest Lecture (Mar 6). Energy-aware hardware, Nergis Tömen. in-person |
12 | 4 | Lecture. Approximate Computing for Green Software. in-person |
 | 5 | Project 2 - steering meeting 1. hybrid |
 | 6 | Project 2 - steering meeting 2. hybrid |
 | 7 | Project 2 - steering meeting 3. online |
19 | 7 | Deadline project 2 – March 28. |
20 | 8 | Presentation project 2 – April 4. |
Assignments
Below the description of each project.
🛠Project 1 – Measuring Software Energy Consumption
- Goal: Compare energy consumption in common software use cases.
- Examples:
- Different versions of the same app;
- Same use case but different apps
- Same version, same app, but different user settings (e.g., enable/disable GPU optimisation)
- Same version, same app, but different running environment
- Different versions of the same app;
- Examples:
- Blog-style report (markdown, approx 2500 words).
- Bonus if you can automate the experiment and there is a replication package.
- Submission by pull request to the website. Instructions here.
- Blog-style report (markdown, approx 2500 words).
- Replication package.
- Weight in final grade: 40%
- Deadline Friday, Mar 1
🛠Project 2 – Hacking Sustainability
- Goal: Solve a Sustainable Software Engineering problem.
- Identify/Describe 1 problem that should be fixed to help enabling sustainability in the software engineering industry/community. You can reuse ideas from the slides.
- Propose a solution. It can be a tool, framework, guidelines, etc.
- Implement the solution.
- Validation. (Depending on the idea)
- Dissemination/social impact. (E.g., solution should be open source, welcome contributors, post on social media, tool website, available in package managers, etc.)
- Deliverables.
- Paper-like article. (Min 4 pages, max 10)
- Online git repo with open source codebase and/or replication package.
- Presentation: 5 min + 5min Q&A
- Weight in final grade: 60%
- Steering meeting/formative assessments: Every week from week 5 to 9.
- Deadline Thursday, March 28.
- Presentation Thursday, April 4.
- Submission by pull request to the website. Instructions here.
Further reading (optional)
- Cruz, L., de Bekker, P. (2023) All you need to know about Energy Metrics in Software Engineering.
- Wohlin, C., Runeson, P., Höst, M., Ohlsson, M. C., Regnell, B., & Wesslén, A. (2012). Experimentation in software engineering. Springer Science & Business Media.
- Cruz, L., & Abreu, R. (2017, May). Performance-based guidelines for energy efficient mobile applications. In 2017 IEEE/ACM 4th International Conference on Mobile Software Engineering and Systems (MOBILESoft) (pp. 46-57). IEEE.
- Cruz, L., & Abreu, R. (2019). Catalog of energy patterns for mobile applications. Empirical Software Engineering, 24(4), 2209-2235.
- Stol, K. J., & Fitzgerald, B. (2018). The ABC of software engineering research. ACM Transactions on Software Engineering and Methodology (TOSEM), 27(3), 1-51.
- Spinellis, Diomidis (2017). The Social Responsibility of Software Development. IEEE Software.
Interesting pointers
- EnergiBridge - Tool to measure energy consumption from almost any device.
- Principles of Sustainable Software Engineering
- CAT community
- Energy Patterns for Mobile Apps.
- Branch Magazine.
- Green Software Lab
- Website Carbon Calculator
- Carbon Tracker
- Greenpeace — Europe’s Green Recovery As if the planet mattered
- Green TU