University of Oulu Sustainable and Autonomous Systems Master's Thesis Guide: 521976S, 30 ECTS, Seminar, Maturity Test and Laturi

Quick answer: what SAS students need to know

For the University of Oulu home path of Sustainable and Autonomous Systems, the current 2026-2027 programme object is 51721 / IMP2026SAS. The exact thesis is 521976S Master’s Thesis in Electronics and Communications Engineering, worth 30 ECTS. The thesis module also contains 521362S Electronics and Communications Engineering Seminar, currently 0 ECTS, and 521011S Maturity Test for Master’s Degree, Electronics and Communications Engineering, also 0 ECTS. The practical thesis process follows the current Electronics and Communications Engineering guidance that explicitly includes Sustainable and Autonomous Systems.

1. What this guide covers

This guide covers the University of Oulu home path of the international Sustainable and Autonomous Systems programme. The programme is a collaboration between the University of Oulu and the University of Vaasa, but a student applies to one university and follows the home-university structure that applies to that admission path. The thesis details here therefore use the current Oulu programme object, Oulu thesis object, Oulu ECE thesis workflow, Laturi, the current seminar and maturity-test objects, and Oulu graduation guidance. It does not attempt to describe the Vaasa home path.

2. Programme identity and degree

The Oulu programme page describes Sustainable and Autonomous Systems as a two-year Master of Science (Technology) programme worth 120 ECTS. The current Oulu study-guide object is 51721, programme code IMP2026SAS, for the 2026-2027 structure used in this guide. The programme combines computer science, electronics, artificial intelligence, machine learning, embedded and real-time systems, wireless technologies and sustainability. Students can encounter themes such as robotics, intelligent machines, smart energy, connected vehicles, environmental sensing, logistics, positioning and cybersecurity.

3. Oulu and Vaasa collaboration boundary

The programme combines complementary strengths of the University of Oulu and the University of Vaasa and can include cross-university study opportunities. That collaboration is academically important, but it does not mean that all thesis administration is interchangeable between the two universities. This guide intentionally follows the University of Oulu home path. When a requirement affects registration, thesis objects, Laturi, examination, maturity or graduation, use the rule attached to the student’s actual home university and current study plan rather than assuming a joint-programme label creates one universal procedure.

4. Exact thesis object

The exact current Oulu thesis object for this programme is 521976S Master’s Thesis in Electronics and Communications Engineering, course object 11669, worth 30 ECTS. It is an Advanced Studies course assigned to Electrical Engineering. The current implementation listed for the 2026-2027 period is 521976S-3002. The course uses the 1-5/FAIL scale and lists Finnish and English as teaching languages. The current course object lists Marko Neitola as the person in charge, but students should still follow the programme’s actual supervision allocation rather than treating the course contact as their automatic thesis supervisor.

5. What the 30 ECTS thesis is expected to demonstrate

The thesis learning outcomes emphasise more than producing a technical artefact. The student is expected to apply creative thinking and problem solving to develop knowledge or procedures, use methods from the relevant discipline, recognise strengths and limitations of methods, and manage the work independently. In a Sustainable and Autonomous Systems thesis, this normally means that a prototype, model, sensing system, software implementation or simulation must be connected to a defined research or engineering problem, an appropriate method, traceable evidence and a critical discussion of limitations.

6. Thesis and Related Studies module

Programme 51721 currently contains a 30 ECTS Master’s Thesis and Related Studies module. The module includes 521976S thesis, 521011S maturity test and 521362S seminar. The seminar and maturity test carry zero credits in their current course objects, but their presence in the programme structure matters. A student should therefore not interpret “0 ECTS” as “irrelevant” or omit them solely because they do not increase the numerical thesis credit total. The correct way to read the module is 30 ECTS of thesis work plus the required zero-credit completion components attached to the current path.

7. Seminar object

The exact current seminar is 521362S Electronics and Communications Engineering Seminar, object 10982. It is an Advanced Studies course assessed PASS/FAIL and currently carries 0 ECTS in the SAS thesis module. Its current implementation is 521362S-3008. The course requires the student to prepare and give a thesis presentation, with the current description specifying about 30 minutes including questions and discussion. Presentations may be given in English or Finnish, and seminars are arranged throughout the year when necessary.

8. The optional one-credit seminar detail

The seminar course description also says that participation in three additional seminars can produce one credit unit that may be included in optional studies. This must not be confused with the core SAS thesis-module seminar. The exact programme structure currently contains 521362S as a 0 ECTS component of the thesis module. The possible one-credit outcome is a separate optional-study consequence of additional seminar participation, not evidence that the master’s thesis itself becomes 31 ECTS or that the required seminar should be counted as extra thesis credit.

9. Maturity-test object

The exact current maturity course is 521011S Maturity Test for Master’s Degree, Electronics and Communications Engineering, object 4157, worth 0 ECTS and assessed PASS/FAIL. The current implementation is 521011S-3006. The course says the maturity test is evaluated and approved by the thesis supervisor and may be taken when the thesis is complete or in its finishing stage. The course description refers to a controlled written event on a topic provided by the thesis supervisor, with an indicative length of approximately three pages.

10. Current maturity delivery workflow

The current ECE thesis process that explicitly covers Sustainable and Autonomous Systems gives a more operational delivery description: the supervisor creates an E-exam, the student completes the maturity test in Examinarium, and Peppi registration is also required. These statements are not treated as contradictory in this guide. The course object provides the content and assessment boundary, while the current ECE process provides the delivery workflow. Because maturity-language requirements depend on the student’s earlier education and University rules, the language labels in the course page are not a universal free-choice rule.

11. Finding a thesis topic

Current ECE guidance says thesis topics may come from companies, research institutes, the university or the student’s own proposal. For SAS students this creates a wide field: autonomous machines, sensing, energy systems, communications, software-defined systems, AI-enabled products, industrial monitoring, smart mobility and sustainability-oriented electronics can all be plausible contexts. The useful question is not whether a topic sounds modern, but whether it can be turned into a bounded research or engineering question with suitable data, measurements, simulation, experiments or evaluation within a 30 ECTS master’s thesis.

12. Contacting the right academic person

After finding a possible topic, the current process directs the student to contact the person responsible for the teaching field most closely related to the work. That step matters in an interdisciplinary programme because SAS topics can sit between electronics, communications, sensing, AI, software and energy. A good early contact should help establish whether the topic belongs within the programme, whether the planned method is academically suitable, and who can supervise it. Do not choose a supervisor solely because a name appears on a course page if the topic belongs to another teaching field.

13. Main supervisor requirements

The current ECE process describes the proposed supervisor as full-time University teaching or research staff with a Doctor of Technology and specified previous thesis-examiner or supervisor experience. The practical point is that supervision is a formal academic role, not merely technical mentoring. A company engineer or research collaborator may be essential to the project, but the University supervisor remains responsible for the academic process. Students should confirm the approved supervisory arrangement before investing heavily in data collection, implementation or experiments.

14. Technical supervisor for external work

If the thesis is carried out outside the University, current ECE guidance says a technical supervisor is always appointed. A technical supervisor may also be appointed in other cases when needed. This role can be particularly important for industrial SAS work involving proprietary hardware, embedded platforms, robotics, production systems, communications infrastructure or energy devices. The technical supervisor can support domain-specific execution, but the academic thesis still needs a defensible research problem, documented methods, evidence, evaluation and academic discussion under the University’s supervision and examination process.

15. Kick-off meeting

The kick-off discussion is where the project should become operational. Current guidance says it defines the topic, possible technical supervisor, completion schedule, supervision implementation and assessment criteria. Students should use this stage to clarify what evidence will count as success. For example, a prototype thesis should not end with “the system works”; it should define measurable performance or evaluation criteria. A machine-learning thesis should decide what training, validation and test evidence is required. A sensing thesis should define calibration and uncertainty. A sustainability thesis should specify the sustainability dimension being measured.

16. Starting the thesis in Laturi

The student starts the thesis in Laturi and invites the main supervisor and any other supervisors. Current ECE instructions also place responsibility on the main supervisor for inviting thesis examiners. Laturi is therefore not merely a final-upload website. It is part of the formal thesis workflow, including personnel, research-plan handling, plagiarism-check related steps, submission and final grading. Students should make sure the Laturi project reflects the actual approved supervision arrangement and should not leave system registration until the end of the research.

17. Research plan

After the supervisor accepts and programme personnel are confirmed, the research plan can be submitted in Laturi and must be approved by the supervisor. A strong SAS research plan should connect four things: the research or engineering question, the proposed method, the evidence that will answer the question, and the limits of the proposed design. For a hardware project that may mean test conditions and reference measurements. For AI it may mean datasets and validation strategy. For autonomous systems it may mean sensors, environmental assumptions and failure scenarios. For sustainability it may mean an explicit metric rather than a broad claim.

18. Sustainable autonomous systems as a method environment

The programme’s method environment is broad. Current course evidence covers sustainable autonomous systems, IoT, machine vision, sensing and tracking, energy harvesting, security engineering, electronic-system design, sensors and measurement, printed electronics, telecommunications, battery testing, intelligent software, software-defined systems, data mining, machine learning, deep learning, multi-modal data fusion, image processing and digital filtering. These courses do not dictate one thesis method. Instead, they show the kinds of technical and analytical tools that can support a programme-relevant thesis when the selected method matches the actual research question.

19. IoT and end-to-end systems

The IoT course evidence covers end-to-end software pipelines, networking, sensor technologies and sensing-data analytics. A thesis in this area should therefore describe the complete chain relevant to the claim: where data originate, how they are sampled, transmitted and stored, what preprocessing occurs, and how outputs are evaluated. If the research claim concerns latency, reliability, energy consumption or sensing quality, the thesis should measure the corresponding part of the chain. A cloud dashboard alone does not validate a sensing method, and a sensor reading alone does not validate an end-to-end IoT system.

20. Machine vision

Machine Vision covers image acquisition, feature extraction, motion, 2D and 3D geometry, recognition and deep-learning fundamentals. A vision thesis should distinguish the imaging conditions from the algorithm. Lighting, viewpoint, optics, sensor resolution, motion, annotation quality and deployment environment can materially affect results. When comparing models, report the evaluation protocol and the conditions under which the numbers were obtained. If a model works only in a controlled laboratory setting, state that boundary instead of presenting the result as general autonomous-system performance.

21. Sensing, tracking and autonomy

Fundamentals of Sensing, Tracking and Autonomy covers sensor noise, calibration, uncertainty, stochastic models, sensor networks, filtering, localisation and tracking. These are especially important when a thesis makes a claim about autonomous behaviour. The system’s decision may be downstream of imperfect sensing, so evaluation should not hide sensor uncertainty. Where relevant, document calibration procedures, reference measurements, localisation assumptions, update rates, environmental conditions and failure cases. A navigation result is stronger when the thesis explains not only the final trajectory but also the uncertainty in the information used to produce it.

22. Energy harvesting and sustainable power

Energy Harvesting Technologies covers solar, kinetic and thermal harvesting, storage, hybrid sources, wireless-sensor-network power and sustainability dimensions. For a thesis involving energy autonomy, the evaluation should connect harvested energy to actual load demand. Useful evidence can include power profiles, conversion efficiency, storage behaviour, duty cycle, environmental conditions and long-term energy balance. A system is not sustainable merely because it uses an energy-harvesting component. The thesis should define what sustainability or autonomy means in the particular study and show how the chosen measurements support that claim.

23. Battery fabrication and testing

Battery Fabrication and Testing covers battery architectures and chemistries, fabrication, electrochemical testing, impedance methods and test-protocol design. A battery-related SAS thesis should therefore be explicit about chemistry, cell format, cycling conditions, temperature, current or voltage limits, instrumentation and analysis methods. When comparing devices, keep operating conditions comparable. If degradation, state of health or energy performance is the research focus, the protocol should be planned before data collection. The test method is part of the evidence, not a procedural detail that can be reconstructed after the experiment.

24. Security engineering

Security Engineering requires threat-model analysis and identification of weaknesses in complex systems. A security claim should therefore state the asset being protected, the assumed attacker capabilities, trust boundaries, relevant attack surface and what the evaluation demonstrates. Saying that encryption is present is not equivalent to showing that a system is secure. For autonomous or connected devices, security can interact with sensing, communications, software updates, authentication and physical access. The thesis should keep the security claim within the tested threat model and explicitly identify important risks that remain outside it.

25. Electronic system design

Electronic System Design covers power supply, thermal design, grounding, transmission lines, crosstalk and complete device or system design. A hardware thesis should connect these design choices to the measured outcome. If the claim concerns signal integrity, thermal performance or reliability, include the corresponding measurements or simulations. If a board or prototype fails under some conditions, that failure can be useful evidence when analysed. The thesis should explain design trade-offs and why the selected architecture was reasonable for the defined requirements rather than presenting the final schematic as self-evident proof of engineering quality.

26. Sensors and measurement systems

Electronic Sensors and Measurement Systems cover sensor selection, accuracy, uncertainty, signal conditioning, multisensor systems, data acquisition and data transmission. A measurement-focused thesis should document the entire acquisition chain relevant to the conclusion. State the sensor range and resolution, calibration or reference method, sampling arrangement, conditioning, synchronisation if multiple sensors are used, and the uncertainty that matters for the claim. If the system is evaluated against a reference instrument, describe the comparison method. A large dataset does not compensate for an unexamined measurement chain.

27. Telecommunications

Telecommunication Engineering covers channel effects, noise, interference, modulation performance, reliability and system simulation. A communications thesis should state the assumed channel, interference model, bandwidth, modulation or protocol conditions and the metric being compared. Simulated and measured results should be distinguished. When claiming reliability or improved performance, report how the result changes across relevant operating conditions rather than only at the most favourable point. Autonomous systems often depend on communication availability, so the limitations of the communications link may also be part of the system-level discussion.

28. AI, data mining and validation

Towards Data Mining, Machine Learning and Deep Learning provide a clear evaluation boundary for AI-oriented theses. Data collection and source combination should be documented, missing or incorrect values handled explicitly, and privacy considered where relevant. Training, validation and test evidence should be separated. Model selection should not be performed on the final test set. Report class balance, baseline comparisons and suitable metrics. Complex architectures such as CNNs, transformers, GANs, VAEs or diffusion models do not remove the need for a defensible validation design and an honest account of generalisation limits.

29. Multi-modal data fusion

Multi-Modal Data Fusion covers alignment, Bayesian inference, parameter estimation, machine learning and multi-sensor fusion. A fusion thesis should explain how different streams are aligned in time and space, how missing or delayed data are handled, and what assumptions are made about sensor dependence and uncertainty. If one modality dominates performance, say so. If fusion improves a metric, compare it against suitable single-modality baselines. In autonomous systems, synchronization errors can look like model errors, so the data-integration procedure should be part of the research method rather than hidden in implementation details.

30. Image and signal processing

Digital Image Processing covers enhancement, restoration, compression, morphology and segmentation, while Digital Filters covers sampling, the DFT, the Z-transform, filter design, finite-word-length effects and multirate processing. These methods can support sensing and autonomy, but preprocessing choices should be justified. Filtering can suppress noise and also remove relevant information. Compression can reduce bandwidth and affect downstream accuracy. A thesis should show why the selected processing chain is appropriate and, where possible, examine sensitivity to important parameter choices instead of treating preprocessing as a neutral step.

31. Sustainability claims need a defined metric

The programme explicitly includes sustainability, but a thesis should not infer sustainability from the word “autonomous” or from the use of AI. Define the dimension being studied. It might be energy consumption, harvested-energy balance, battery lifetime, material use, device lifetime, maintenance demand, emissions, resource efficiency or another defensible indicator. Then connect the method to that indicator. A technical improvement can support a sustainability argument only when the causal link is explained and the evidence is appropriate. Keep broader environmental or societal conclusions within what the study actually measures.

32. Research data management

University responsible-research guidance applies to SAS theses just as it does to other research. Plan how data will be collected, named, stored, backed up, processed, documented and retained. For experiments, keep enough metadata to reproduce conditions. For simulations, record model versions and parameters. For ML, document dataset versions and preprocessing. For industrial projects, clarify what can be published and what must remain protected. The general principle is to make outputs as open as possible while restricting material that genuinely needs protection under contractual, privacy, security or other legitimate requirements.

33. Personal data and privacy

Autonomous-system research can involve personal data even when the project is primarily technical. Images of people, voice, location traces, vehicle data, device identifiers, user logs and interaction data may make individuals identifiable. Follow current University privacy guidance when personal data are processed, including minimisation and appropriate security. Pseudonymisation reduces direct identifiability but does not automatically make data anonymous. The thesis should describe the data-processing boundary accurately and avoid claiming that a dataset is anonymous merely because names were removed.

34. Ethics review

Ethics-committee review is not automatic for every SAS thesis. Whether review is required depends on the actual human-research design and the current criteria of the relevant ethics committee. A purely technical simulation or bench measurement will not have the same ethical profile as a study involving human participants, behavioural monitoring or sensitive personal data. Students should assess the need for ethical review early because approval cannot simply be added after data collection. The guide therefore avoids both extremes: it does not say every SAS thesis needs ethics approval, and it does not treat technical research as automatically outside ethics considerations.

35. AI use and research integrity

University ethical principles apply to sources, authorship, data, code and misconduct prevention. AI-assisted work does not transfer responsibility away from the student. Generated code must be tested; generated references must be checked against real sources; generated summaries must be compared with the original material; and generated analysis must not be presented as empirical evidence unless it actually follows the research method. For a thesis involving AI as the research object, distinguish clearly between using an AI tool to assist the writing process and evaluating an AI model as part of the scientific work.

36. When the thesis is ready for assessment

Current ECE guidance says the student receives permission to upload the thesis to Laturi when the supervisor considers it ready for assessment. This creates a useful operational boundary: final submission should follow supervisor approval rather than simply the student’s preferred deadline. Before that point, check that the research question, method, results and conclusions align; figures and tables are traceable; limitations are explicit; required seminar and maturity arrangements are understood; and confidential information has been handled correctly. The final version should be academically complete before it enters the formal examination process.

37. Examination workflow

The current process says the programme responsible person and examiners download the thesis and inspect the plagiarism-check result in Laturi. Examiners then prepare an evaluation proposal using the programme evaluation form, with the proposal due three days before the Degree Programme Committee meeting. The Committee evaluates the thesis based on the examiner statements and updates the final grade in Laturi. This is why students should not assume that one supervisor alone decides the final grade. The current public evidence uses plural “examiners” but does not establish a universal fixed examiner count for every SAS case.

38. Assessment and correction route

The thesis object uses the 1-5/FAIL scale. The current ECE process also provides a 14-day correction route after notification of the assessment decision. Students should distinguish correction of the assessment decision from ordinary thesis revision before grading. The latter happens during supervision; the former is part of the formal post-decision process. Keep important supervision and submission records, especially if the work involves external partners or unusual evidence. If a disagreement arises, rely on the current official assessment and rectification instructions rather than informal assumptions about who can change a final grade.

39. Publication and confidentiality

Company-funded or industrial theses can involve confidential information, but confidentiality should be handled deliberately. The public thesis should not contain protected material that the student is not allowed to publish. At the same time, the academic work still needs enough method, evidence and explanation to support the conclusions. A practical solution may require separating public thesis content from confidential background material, datasets or implementation details, depending on the agreement and University rules. Resolve these boundaries early, because removing essential evidence at the final stage can weaken the academic argument.

40. Graduation

After the thesis, maturity test and other degree requirements are complete, the degree-certificate application proceeds through Peppi under current University graduation guidance. Students should verify their study record before applying, especially the thesis grade, maturity result, seminar completion and any cross-university studies that must appear correctly in the record. The joint Oulu-Vaasa nature of the programme does not remove the need to complete the home-university graduation process. If a cross-university course is missing or recorded incorrectly, resolve that issue before assuming the degree application will automatically reconcile it.

41. A realistic thesis sequence

A practical sequence is: identify a programme-relevant topic; confirm the responsible teaching field; establish the University supervisor and any technical supervisor; hold the kick-off meeting; define question, method and evidence; open the thesis in Laturi; submit and obtain approval for the research plan; complete data collection, simulation, implementation or experiments; analyse the evidence; write and revise the thesis; complete the thesis seminar; arrange the maturity test through the current ECE/Examinarium workflow; obtain permission for final Laturi upload; complete examination; and apply for graduation after the required records are present.

42. Common mistakes

Common mistakes include using the standalone Electronics thesis code 521977S instead of the SAS thesis 521976S; importing the CSE thesis 521993S; ignoring the zero-credit 521362S seminar; assuming the course-page language list gives every student free choice for the maturity language; starting industrial data collection before privacy, confidentiality or supervision boundaries are clear; reporting only best-case autonomous-system performance; mixing training and test data in ML evaluation; calling a system sustainable without a measured sustainability indicator; and assuming that a sophisticated prototype is automatically a strong thesis without explicit research reasoning.

43. What not to import from other Oulu programmes

Several nearby Oulu programmes share technical fields or administrative systems, but their thesis objects are not interchangeable. Standalone Electronics uses a different exact thesis object, and standalone CSE uses another. SAS currently shares the ECE seminar and maturity objects, but its own programme structure points to 521976S for the thesis. Keep the distinction at the object level. Similar course titles, shared staff, Laturi usage or common faculty guidance do not justify replacing the exact programme-specific thesis code with a code from another study option.

44. Freshness and the 2027-2030 curriculum transition

This guide is bounded to the 2026-2027 programme evidence currently attached to object 51721 / IMP2026SAS. The University has published curriculum-transition information for 2027-2030, so students starting or continuing under the autumn-2027 structure should not assume that every present object, implementation code or administrative detail will remain unchanged. The durable rule is to recheck the current programme structure, thesis object, seminar, maturity route and ECE process before treating this guide’s 2026-2027 implementation details as final for the new curriculum period.

45. Practical final checklist

Before final submission, confirm that your Peppi structure points to 521976S, not another programme’s thesis code; your Laturi project has the correct supervision personnel; your approved research plan matches the work actually completed; your method supports the research claim; autonomous-system evaluation reports relevant uncertainty and failure conditions; ML evaluation separates training, validation and test evidence where applicable; sustainability claims have a defined indicator; privacy, confidentiality and ethics boundaries are resolved; the 521362S seminar requirement is completed; the 521011S maturity route is completed; and final graduation records are correct in Peppi.