Sustainable Software Engineering 🌱
CS4415 | Edition of 2021/22
| Delft University of Technology | [MSc in Computer Science – Software 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 economic 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 will cover a set of competencies needed to leverage sustainable software systems. It will have a strong component on Green SE, covering techniques to measure and improve the energy-efficiency at any stage of the software lifecycle. Students will learn about 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 🔒 | https://brightspace.tudelft.nl/d2l/home/399685 |
| Instructors | Luís Cruz, [Arie van Deursen] |
| Schedule | Tuesdays 10:45, Thursdays 8:45, Fridays 10:45. 🔗 MyTimetable. |
| Mattermost | 🔗 Sign-up link |
| ECTS | 5.0 |
| Quarter | Q3 |
| Format | Classes are optimised for in-person attendance; Online attendance is allowed in most classes. |
| Examination type | Project 1 - Essay (30%); Project 2 - Software Repository and Essay (60%); Presentations (15%). |
| 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. 📊Slides 🎥Recording hybrid |
| 2 | 1 | Lecture. Social and Individual Sustainability in SE. hybrid 📊Slides |
| 3 | 1 | Paper Reading. Read and summarise research literature on Social and Individual SE Sustainability. |
| 4 | 2 | Presentation. Present a summary of a research paper. hybrid |
| 5 | 2 | Lab. Measuring software energy consumption. Introduction to Project 1. 📊Slides 🎥Recording hybrid |
| 6 | 2 | Lecture. Green Software Engineering — Part I: units of energy. 📊Slides 🎥Recording hybrid |
| 7 | 3 | Project. Project 1 - steering meeting and formative assessment. online/in-person |
| 8 | 3 | Lecture. Green Software Engineering — Part II: Scientific guide for reliable energy measurements. 📊Slides 🎥Recording part2 hybrid |
| 9 | 3 | Lecture. Green Software Engineering — Part III: Energy efficiency in mobile computing; carbon-aware data centres. 📊Slides 🎥Recording hybrid |
| 10 | 4 | Project. Project 1 - steering meeting. online/in-person |
| 11 | 4 | Lecture. Green AI - 📊Slides. MSc Talk by Tim Yarally - 📊Slides. 🎥Recording hybrid |
| 12 | 4 | Guest Lecture. Saving energy in software development by making the right choices. Stefanos Georgiou. 📊Slides 🎥Recording online |
| 13 | 5 | Lecture. |
| 14 | 5 | Guest Lecture. The effect of design patterns on energy consumption. Daniel Feitosa. 📊Slides 🎥Recording in-person |
| 15 | 5 | Project. Project 2 - Description and kick-off. 📊Slides hybrid |
| 16 | 6 | Steering Meeting. online |
| 17 | 7 | Steering Meeting. hybrid |
| 18 | 8 | Steering Meeting. Formative assessment. hybrid |
| 19 | 9 | Presentation and Demos in-person |
Assignments
Below the description of each project.
📚 Literature Review
- Groups of 2 or 3
- Select 1 academic paper that talks about social or individual sustainability in the context of software engineering.
- Google Scholar, DBLP
- Double check with the lecturer whether your selected paper is good enough (use mattermost).
- Write a short summary about it (min 200 words; max 500).
- Prepare a presentation for next Tuesday (Feb 15).
- 7 min + 3 min Q&A.
- Presentation is 7.5% of the grade. (But you can recover in the presentation of project 2)
- Submission by pull request to the website. Instructions here.
- Deadline Tuesday, Feb 15.
🛠 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. ⭐️(new)
- Weight in final grade: 30%
- Steering meeting/formative assessment: Tuesday, Feb 22
- Deadline Thursday, Mar 3
🛠 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.
- Propose a solution. It can be a tool, framework, guidelines, etc.
- Implement the solution.
- Validation. (Depending on the idea) (side note: the cancelled class was all about this)
- Dissemination/social impact. (Solution should be open source, welcome contributors, post on twitter, hacker news, reddit? Tool website?)
- Deliverables.
- Paper-like article. (Min 4 pages, max 10)
- Online git repo with open source codebase and/or replication package.
- Presentation: 7 min + 5min Q&A
- Weight in final grade: 60%
- Steering meeting/formative assessments: Every week from week 5 to 9.
- Deadline Tuesday, April 5. (Grace period until April 8).
Further reading (optional)
- 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.