The course consists of two modules, Sensing and Perception and Control and Decision making. The introduction to the course is given at the Kick-off in January. After this, there is a preparation phase where you read up on the topics of the first module and do some preparatory assignments. During this phase you will work on your own and have meetings with students from your local university. It is important that time is allocated for this. After this follows a two day meeting where we have presentations related to the topics and hands-on sessions to help set up for the examination that follows in the third step. These three steps are then repeated for the two last topics.
The WASP courses play an important role in the WASP graduate school beyond providing a source of knowledge and skills in autonomous systems and software. They provide an arena where people meet. They are vital in the process of forming a network between industry and academic and between different universities and between different companies. During the meetings you will meet your colleagues in the WASP program. You are part of the second batch of more than 40 students. In total there are >100 PhD students in the WASP program. This makes WASP one of the largest research programs ever in Sweden. Therefore, look at the course as an opportunity to meet and get to know people. Interact with students also outside your own area of research and from different parts of Swedish industry.
- Session 0: A short introduction to the course and to autonomous systems
- Lund Jan 8, as part of the kick-off
- Session 1: Sensing and Perception
- Linköping Feb 22-23, roughly 10:30 first day to 15:00 second day
- Session 2: Control and Decision Making
- KTH May 3-4, roughly 10:30 first day to 15:00 second day
A rough overall schedule looks like this:
- Jan 8 Introduction
- Jan 11 – Feb 21 Preparation work for topics Sensing and Perception with local sessions
- Feb 22 – Feb 23 Two day meeting in Linköping
- Feb 26 – Mar 23 Local examination activities with local sessions
- March 29 Deadline for examination assignments in module 1
- Apr 2 – May 2 Preparation work for topics control and decision making with local sessions
- May 3 – May 4 Two day meeting in Stockholm
- May 7 – June 1 Local examination activities with local sessions
- June 7 Deadline for examination assignments in module 2
Teaser about the content
So what will the course be about you probably ask. What does the sensor data from the typical sensors used in, for example, autonomous cars today look like and what information can you get out of them? Deep learning is said to solve one problem after the other. What is really behind this and what tools are there to use? You might have taken a course on control, but how is control design done in a complex system with many components and how to evaluate if it is time to look at redesigning the system? What if the decisions you need to make are not about how to control a certain signal but rather achieving an objective for an entire system or several system with many different abilities. Clearly you need some abstraction, but how do you formulate such a planning problem and how do you solve it? These are some of the things that you will work with in the course. Sounds exciting? We think so!
PhD students: The course will take time, you should expect to spend at least 4 weeks of full time work on the course. You need to allocate this time.
Supervisors: It is important that sufficient time is allocated for the students to participate in the course activities. The students’ study plans should reflect this.
Team behind the course
The development of the new course is lead by the Graduate School Management (GSM) group. Each of the participating universities have one representative in the GSM. The representatives are Bo Bernhardsson (LTH), Christian Berger (Chalmers/GU), Fredrik Heintz (LIU) and Patric Jensfelt (KTH). These people will lead the local sessions during the course and will be the point of contact at the corresponding university during the course.
Changes from previous course round
When talking to WASP students from the first batch it is good to keep in mind that there has been some restructuring in the course based on the experience from the first course round. This course corresponds to the first half of the 12hp Autonomous Systems course that the students from the first batch attended. The new 6hp course runs over one semester and so did the previous 12hp full course. This means that there is twice the calendar time for the course content compared to the first course round.
In the previous course, each module was initiated with a two day meeting with lectures on the topics from the module. After this followed the work on the examination tasks. The examination was composed of individual exercises, lab assignments done in pairs and a project assignment done in groups of 5-6 people. The exercise were of theoretical nature and so were some of the lab assignments. The project was very hands on and stretched from the start to the end of the course, gradually adding functionality connected to the topics as they were covered in the course.
The upcoming course will focus more on tools that can be used in the different domains. This means that the examination activities will be more hands on and this will replace the project work. Instead of having traditional lectures, which are hard to adapt to such a diverse crowd, you will study material more on your own, using both written and video material. To ensure that the students of the course work actively with the material in the preparation there will be some introductory tasks during the preparation phase which will lead up to the examination assignments.