AI Planning for Ubiquitous Computing

Hauptseminar

General

Lecturer

Dr. Ilche Georgievski

Study program

Computer Science, Software engineering, Information Technology

Prerequisites

Previous knowledge of the course "Smart Cities and Internet of Things" is helpful but not necessary.

Events

Kick-off meeting

The Kick-off meeting will take place on 9.11.2020 from 13:00-15:00 in the following WebEx room: https://unistuttgart.webex.com/meet/ilche.georgievski

Withdrawal

16.11.2020

Block event

To be announced.

Attendance

Students have to be present at the kick-off meeting and the block event.

Description

Ubiquitous computing aims at creating ambience in which one's experiences and quality of life are improved by monitoring and assisting people using the Internet of Things and intelligence in coherence. The current level of intelligence in ubiquitous computing environments is achieved by using predefined sequences of actions that are usually executed in conditions determined in advance and for well-known situations. Such level of intelligence is limited in that it cannot deal with the dynamics and uncertainness of ubiquitous computing environments, it does not consider the needs of those populating the environments, and it is ignorant of some global objectives, such as energy saving. Ubiquitous computing is therefore in need of techniques that go beyond predefined solutions, and act automatically with sophisticated intelligence. The field of Artificial Intelligence (AI) focuses on developing highly flexible and effective systems for intelligent behaviour, where AI planning provides a means for automated and dynamic creation of plans. The field of AI planning was initiated in the 1970's and has made significant progress in theory and practice since. AI planning provides powerful techniques that require a goal, an initial state of an environment, and some knowledge about the environment to select and combine a course of action that, when executed in the initial state, achieves the goal. This makes AI planning highly suitable for ubiquitous computing. In this advanced seminar, we will study works that propose AI planning approaches for ubiquitous computing environments.

Organisation

Seminar topics will be introduced during the kick-off meeting. Students should select three topics in preference order and inform us. Topics will be distributed based on priority while considering the topic preferences. Students are requested to research the topic, write a paper, and prepare an oral presentation.

Paper

Students have to hand in a written paper about the chosen topic. The paper should be prepared using LaTeX.

Page limitation: maximum 15 pages.

Template: To be provided.

Instructions: To be provided.

Deadline: Students must submit the paper two weeks after the block event.

Presentation

Students have to give an oral presentation about the chosen topic.

Duration: 25 minutes + 5 minutes for Q&A

Template: To be provided.

First deadline: Students must submit the first draft of the presentation to the supervisor one week before the block event.

Second deadline: Students must submit the final presentation to the supervisor on the day of the block event.

Grading

The presentations, papers, and active participation will be graded.

Language

Papers and presentations must be in English.

Vorlagen

Ausarbeitung

Präsentation

Richtlinien, Hinweise, Best Practices und Literaturhinweise

Mehr zum Thema Seminar- und Hauptseminararbeiten finden Sie hier:

Zum Seitenanfang