Course syllabus

Course-PM

SSY345 Sensor fusion and nonlinear filtering lp4 VT24 (7.5 hp)

The course is offered by the Department of Electrical Engineering

Contact details

Examiner
Lars Hammarstrand, lars.hammarstrand@chalmers.se

Teaching assistants
Ji Lan, ji.lan@chalmers.se
Silan Karadag silan@chalmers.se

Student representatives

The purpose of this course is to give a thorough introduction to sensor fusion in time-varying settings (also known as filtering or smoothing), i.e., how to perform state estimation using a variety of sensors. Such methods continue to receive considerable attention due to their high versatility; famous examples include their role in the Moon landing and their current importance in the development of self-driving cars.

In the course, we emphasize the positioning of vehicles, people, mobile phones, robots, etc, though the potential applications go way beyond that. The course provides a solid theoretical background and offers hands-on experience applying the techniques to solve practical problems.

Schedule

The schedule for all activities can be found here. Here you can find all mandatory activities, both in terms of mandatory in-class activities and assignment deadlines, that you need to keep track of.

The in-class activities in the course can be found in TimeEdit.

Note 1: Both practice sessions and solution discussions are half-class. If you are in Practice Session Group 1, you are expected to go to the first occasion, and if you are in Practice Session Group 2, you are expected to go to the second. If you can not attend your given session, you are responsible for finding a fellow student that you can switch with. 

Note 2: The solution discussions last only 45 minutes for each group, so group 1 is expected to come during the first hour and group 2 during the second hour.

Course Literature

We have aimed to make the course self-contained in the sense that you should be able to solve all tasks using the material provided in the course. However, if you would like to read a book covering the material, we recommend the following:

Simo Särkkä and Lennart Svensson, Bayesian Filtering and Smoothing, Second Edition, Cambridge University Press, 2023.

You can find the book for free as a PDF on Simo's homepage here.

Course Design

The course is given in a flipping-the-classroom style, which means that you are expected to watch lecture videos before coming to class, and that the time in class will be used to work more actively with the material in the course. A more detailed explanation of what it is and the motivation for choosing it can be found here.

For the flipped classroom model to be effective, all students must stay up to date with the course material. To ensure this, several mandatory activities and tasks must be completed before their deadlines. The mandatory elements are (click on the links for more information about each activity):

• Online video lectures
  Before each practice session, you are expected to watch a set of online video lectures covering the material of that specific practice session.
• Practice sessions (PS)
  In-class sessions, where we will actively discuss and work with the material of the corresponding video lecture.
• Home assignments (HA)
  After getting familiar with the material through the video lectures and the practice sessions, you will get to apply your knowledge in a set of home assignments that you will solve individually. Each home assignment typically covers the material of two practice sessions. 
• Solution discussions (SD)
  After each home assignment, we hold active sessions to discuss the results, the conclusions, and the insights we have obtained.
• Inspera closed-book exam (INSP)
  As part of the Home Assignment examination, we will test your knowledge regarding the HAs and corresponding lectures with digital closed-book exams that are performed at campus on your own computer.
• Project
  To conclude the course, you get to apply your knowledge with a fellow student to solve a small problem using real measurements.
• Oral examination
  There is no written exam in the course. The examination is done through the home assignments, the project, and a final oral exam. 

Students are allowed 8 “late days” if they cannot make a deadline. These late days do not extend future deadlines. If a late day is used for an in-class practice session (which implies that they miss that session), then the students have to turn in written solutions to problems related to that class (within a week after they receive the tasks). Missing out on a class where we discuss solutions to the home assignments only costs a late day (no need to submit any written solutions). 

Code of conduct

From a pedagogical standpoint, we believe the flipped classroom with continuous examination is an effective way to teach this subject, and course evaluations support this. However, this approach relies on your commitment—preparing for practice sessions by watching the videos and refraining from sharing assignment solutions.

To this end, we have a code of conduct that all students are expected to follow.  

Use of AI tools

AI tools like ChatGPT and MS Co-Pilot are ubiquitous nowadays. While moderate and responsible use of such tools can be helpful, your learning experience will be significantly worse if you rely extensively on AI tools. However, we have no practical way of checking to what extent such tools are used, and have therefore decided on the following sets of rules and recommendations:

• AI tools are allowed for home assignments and the project, but we strongly recommend keeping their usage to a minimum.
• In the same spirit, it's allowed to ask friends, ask TAs, Google stuff, etc., but you learn more from trying yourself until you get really stuck.
• Regardless of your reliance on tools, friends, Google, etc, you need to fully understand your solutions and convince yourself that they are correct.
• For the project, we expect you to draft the report in your own words. After that, you may use AI tools to improve the text, but you still need to convince yourself that the final version is correct and that you personally understand its contents, so that you stand by it and are ready to defend it.

We still believe that the Home Assignments and Project provide a useful learning experience. However, whether you did enough independent work with the home assignments to understand them fully and to earn points for your final grade will be examined through closed-book digital exams. More information about these is provided Inspera closed-book exam (INSP).

Lecture Content

The content of the different lectures can be briefly summarized as follows:

• L0: Course welcome and structure (live). 
• L1: course introduction and a primer in statistics (flipped). 
• L2: Bayesian statistics and decision theory (flipped).
• L3: State-space models and optimal filters (flipped).
• L4: The Kalman filter and its properties (flipped).  
• L5: Motion and measurement models (flipped). 
• L6: Nonlinear Gaussian filters (flipped),
• L7: RTS and Gaussian smoothing (flipped).  
• L8: Particle filters (flipped)
• L9: Guest lecture by Dr. Daniel Svensson, Principal Sensor Fusion Architect, Autonomous Drive at NVIDIA (live over zoom).

See the schedule for when you should have watched and answered the self-assessment questions for the flipped lectures above. Please meet the deadline to avoid losing 1 late day per day. 

Changes this year

There are a few changes made to the course this year:

• Added Inspera closed-book exams as part of the examination of the home assignments. As part of this change, we have transferred points from the analysis report to the  Inspera test. The reason for this is mainly that large language models make it hard to assess student comprehension using home assignments. However, we still believe that making an honest attempt at the home assignments is the best way to learn the content of the course and should be an excellent preparation for the Inspera closed-book exams.
• We have made the peer-review for the home assignments optional this year. You will still be assigned a report after the deadline of the corresponding home assignement but you are not required to provide feedback. If you have the opportunity to do so, we believe that this is very valuable for your fellow students. We also hope that you get insights from seeing how another student has solved the assignments.
• The oral exam at the end of the course will count less towards your final grade. As we have introduced closed-book exams, we see less need for this type of examination. 
• Each year, we strive to make (small) improvements to all parts of the course. 

Learning Objectives and Syllabus

After the course, students should be able to

• explain the fundamental principles in Bayesian estimation
• describe and model the commonly used sensors' measurements
• summarize and compare the most typical motion models in positioning to know when to use them in practical problems
• derive the expression for an optimal filter
• describe the essential properties of the Kalman filter (KF) and apply it to linear state-space models
• implement the key nonlinear filters (above all the extended Kalman filter, unscented Kalman filter, and the particle filter) in Matlab, to solve problems with nonlinear motion and/or sensor models
• select a suitable filter method by analyzing the properties and requirements of an application
• apply the linear and nonlinear Rauch-Tung-Striebel smoothers to general smoothing problems
• solve a variety of important real-world filtering and smoothing problems by employing and adapting the above knowledge to a variety of applications.

Link to the syllabus on Studieportalen.

Study plan

Examination form

There is no written exam in the course. Instead, you will be examined through your performance on the home assignments, the project, and the oral exam.

The possible grades in this course are 3, 4, 5, or not passed. To pass the course and obtain a grade of 3, 4, or 5, you need to fulfill the mandatory parts of the course without exceeding the eight late days. Students who wish to obtain a higher grade can do so by collecting points of excellence (POE) throughout the course, as given in the table below. However, if you fail the oral exam, you will also lose the POEs you have collected and can only get a grade of 3 in the course.

	POE (Max)
Home Assignments (total / HA)	20
Well-illustrated and commented analysis report	6
Inspera closed-book exam	14
Project Report	10
Oral examination (project and course)	10
Total	100

If you submit an analysis or project report late, all POEs over 40% is divided by 2. The same applies to revised solutions (if you also submit them late, you divide by 4).

In total, it is possible to obtain 100 POE, which are mapped to the final grade as

POE	Grade
40-59	3
60-79	4
80-100	5

Note that the performance on quizzes in the lecture videos and in-class practices is not graded apart from mere participation. We do so intentionally to emphasize that we arrange these activities for learning and not as a test. By doing so, we also encourage all participants to ask any questions they find interesting or are struggling with.