Course - Best Practice - Machine Learning for Ship Autonomy - IP505314
IP505314 - Best Practice - Machine Learning for Ship Autonomy
About
Examination arrangement
Examination arrangement: Assignment
Grade: Letters
| Evaluation | Weighting | Duration | Grade deviation | Examination aids |
|---|---|---|---|---|
| Prosjektoppgave | 100/100 |
Course content
In this course, we will introduce variant of machine learning methods and apply them for ship autonomy applications. The aim is to show the potential use of these methods for solving specific problems on autonomous ships, such as path planning, auto-docking and motion prediction. We plan to present case studies for each of introduced machine learning methods.
- Introduction to machine learning (state of the art)
- Dijkstra method, A* method (application: path planning for close-range maneuvering)
- Neural network architecture, including MLP, LSTM and NARX (application: ship motion prediction, and force allocation to thrusters)
- Deep learning method (application: remaining useful life predictions for turbofan engine)
Learning outcome
After course completion, the student should understand the concept of machine learning and autonomy in ship systems, as well as having knowledge of important machine learning algorithms and techniques, specially applied to maritime cases..
The student develops skills in planning, design and applying machine learning techniques to maritime cases.
The student is able to formulate research problems involving machine learning apply its principles in complex systems, such as maritime.
Learning methods and activities
The course is given during two weeks, and is organized with lectures on background topics and an introduction of case studies.
The case study will then be solved individually or in groups and documented in a project report. The total workload for the course is expected to be 2 weeks including independent research and literature survey supporting the project work. Grading will be based on the project report, and will assess the candidate(s) ability to interpret, familiarize, reflect and apply the course topics.
Mandatory assignments:
Mandatory exercises must be approved before admission to the examination.
Compulsory assignments
- Obligatoriske arbeidskrav
Further on evaluation
Project report (100%)
Specific conditions
Admission to a programme of study is required:
Naval Architecture (850MD)
Naval Architecture (850ME)
Product and System Design (840MD)
Product and System Design (845ME)
Recommended previous knowledge
Students must have basic knowledge of machine learning, especially neural networks. Background in Product Design, Automation or Computer Science is also an advantage.
Course materials
Students will test the methods if they are interested. Access to software will be given ahead of the course. Students can download lecture notes from given network.
No
Version: 1
Credits:
3.8 SP
Study level: Second degree level
Term no.: 1
Teaching semester: AUTUMN 2020
Language of instruction: English
Location: Ålesund
- Marine Technology
Department with academic responsibility
Department of Ocean Operations and Civil Engineering
Examination
Examination arrangement: Assignment
- Term Status code Evaluation Weighting Examination aids Date Time Examination system Room *
-
Autumn
ORD
Prosjektoppgave
100/100
Release
2020-12-08Submission
2020-12-14
12:00
INSPERA
12:00 -
Room Building Number of candidates - Spring ORD Prosjektoppgave 100/100 INSPERA
-
Room Building Number of candidates
- * The location (room) for a written examination is published 3 days before examination date. If more than one room is listed, you will find your room at Studentweb.
For more information regarding registration for examination and examination procedures, see "Innsida - Exams"