Home Environment Themes Projects Core Curriculum Organization Events Publications

Upcoming Events

 

See the seminar series at Automatic Control, CMIV, and NVIS/VITA


Past Events


CADICS CCC Course in Planning and Optimal Control 9-16 May 8th 2012
For more information see the course page.



CADICS/SSF Workshop March 1 2012, 10-12 in Visionen
Researchers in CADICS has successfully applied for and received three major grants from the Swedish Foundation for Strategic Research (SSF). To present these projects to the CADICS environment and to discuss the possibilities for synergies between the projects CADICS JF arranges a half-day workshop in Visionen on March 1, 2012.
The projects are:
Collaborative Unmanned Aircraft Systems (CUAS), Patrick Doherty
Cooperative Localization (Coop-Loc), Fredrik Gustafsson
Data-driven Scene Characterization for Realistic Rendering (VPS), Anders Ynnerman

Tentative Program
10.15-10.20 Short introduction
10.20-10-50 Presentation and discussion of the CUAS project
10.50-11.20 Presentation and discussion of the Coop-Loc project
11.20-11.50 Presentation and discussion of the VPS project
11.50-12.00 Summary and general discussion



CADICS Research Workshop 5-6 December 2011
Program

Monday 5/12
12-13 Lunch
13-14 Presentation of CADICS and its new organization
14-16 Presentation of research infrastructure
15-15.30 Coffee
16-18 Group discussions about CADICS course projects
19 Dinner

Tueday 6/12
8.30-9.30 CADICS Core Curriculum
9.30-10 Overview of CADICS Research Themes
10-10.30 Coffee
10.30-12 Group discussion about CADICS research projects
12 Lunch



On August 25-26 (2011) Stephen Boyd is offering a CCC mini-course entitled "Topics in Convex Optimization", click here for more information about the course. Welcome!




Guaranteed Search in Large-Scale Outdoor Terrain

Speaker: Alexander Kleiner
Time and Place
: Friday 2011-03-11 13:15-14:15 in Visionen
Abstract
: In this talk I present a novel approach for guaranteed search in real-world outdoor terrain. Guaranteed search requires a team of searchers to clear an environment of targets moving at unknown speeds. This implies to generate a search strategy that guarantees to locate any moving target in the area, which is generally known to be a NP-hard problem. The problem is relevant, for example, when securing terrain and building structures, such as a museum, but also for large-scale victim search during Search And Rescue (SAR) missions.  The method presented in this talk is based on an elevation map representation of the three-dimensional environment. Elevation maps are particularly suitable for large-scale outdoor scenarios since they can capture 3d visibility of specific sensor types and target heights. From elevation maps a graph representation is extracted, which is used to compute guaranteed strategies for coordinating  a team of searchers (humans or robots) and to generate trajectories based on an appropriate classification of the terrain.  The computational complexity is relaxed in two ways:
First, the graph sampling is optimized by a hierarchical approach that extracts locations with higher visibility first.
Second, schedules are computed in an anytime fashion by considering spanning trees extracted from the graph.
I finally introduce results from experiments with real-world data sets that investigated the performance of the proposed approach. Furthermore, I present results from a large-scale outdoor experiment conducted at the Gascola robot training site nearby Pittsburgh, US.

Short Bio
Alexander Kleiner is "Akademischer Rat" (comparable to assistant professor) at the research group on foundations of Artificial Intelligence in Freiburg. He received his PhD degree from the University of Freiburg in February 2008 where he is lecturing classes in Robotics and Multi-Agent systems since 2001. His research addresses autonomous robot exploration, simultaneous localization and mapping (SLAM), and mixed-initiative teams of humans and multi-robot systems. The main concern of his work is to develop algorithms for the real-world-deployment of robot and multi-agent systems, particularly for supporting humans in hostile situations, such as urban search and rescue (USAR). His work resulted in more than 10 robot and multi-agent systems that won several prizes at international competitions, such as RoboCup and the TechX challenge, and more than 40 peer reviewed scientific papers published in the major scientific journals and proceedings of the field. Since 2006, he is member of the executive committee of RoboCup.

Joint Workshop on Particle Filtering, October 13, 2010



Enhanced Semantic Networks and Action Ontologies for Automated Recommendation Systems Speaker: Leora Morgenstern
Time and Place
: Wednesday 2010-06-16 13:15-14:00 in Visionen
Abstract
: In many business domains, including banking, telephony, and sales, customer service representatives must process large and frequently changing bodies of information in order to recommend actions to their customers. In this talk, I present several knowledge-based technologies that I developed while at IBM that lie at the core of systems for automating these tasks.

First, I present enhanced semantic networks, an extension of standard semantic networks  that supports  provably sound and computationally efficient reasoning. Next, I present SNAP, a business-process and agents-based ontology. I show how these SNAP and ESN were integrated into dynamic dialogue and automated recommendation systems, and discuss the systems that were used for banking and sales applications for Fortune-500 companies. 


Theories of Delegation and Joint Planning in Egalitarian Multi-Agent Environments
Speaker
: Leora Morgenstern
Time and Place
: Tuesday 2010-06-15 13:15-14:00 in Visionen
Abstract
: In this talk, I consider the problem of respectful delegation --- delegation in environments in which all agents have a say in the way their time and resources are used. In such environments, the delegating agent cannot count on delegatees stopping  their activities when they receive requests from the delegator. Nevertheless, the delegator needs to be able to determine when his requests will result in his plan being executed successfully. I present a epistemic temporal theory which supports inferring that a particular request and set of circumstances will result in successful execution of a plan.

I then consider the problem of joint planning, in which a group of agents form a joint intention to achieve some goal, and pool their knowledge to construct a plan in which all agents participate and which achieves the desired goal. I model joint planning as an extension of the problem of respectful delegation, in which the agents mutually form a corporate-like joint plan entity that delegates actions to each of its members.

Opportunistic networks meet infrastructure

Speaker: George Theodorakopoulos
Time and Place
: Tuesday 2010-06-15 9:15-10:00 in Glashuset
Abstract
: Opportunistic networks are composed of mobile wireless devices that communicate when within range of each other, i.e. when there is a contact. No end-to-end connectivity is assumed to exist, so data is stored and carried until an opportunity for forwarding appears. However, since user mobility is slow compared to the data transmission speed, such networks experience higher delivery delay compared to cellular networks. On the other hand, the traffic load of cellular infrastructure networks is increasing due to user demand for content in the Internet.

I will present recent experimental results on the performance enhancements when merging infrastructure with an opportunistic network. From a more theoretical perspective, I will show how a worm propagating in an opportunistic network can spread faster if the infrastructure


Joint CADICS/MOVIII workshop May 18, 2010


Verification and Control of Hybrid Systems using Reachability Analysis
Speaker
: Claire Tomlin
Time and Place
: Wednesday 2010-05-05 13:15-14:15 in Visionen
Abstract
: This talk will present reachability analysis as a tool for model checking and controller synthesis for hybrid systems. We will consider the problem of guaranteeing reachability to a given desired subset of the state space. We allow for nonlinear dynamics in each discrete mode, and possibly non-convex state constraints. We make use of techniques from hybrid system verification to compute reachable sets, under bounded model disturbances that vary continuously. We also consider the effects of sampling and quantization. The resulting control policy is an explicit feedback law involving both a selection of continuous inputs and discrete switching commands at each time instant, based upon measurement of system state. We discuss real time implementations of this, and present several examples from multiple UAV control.


Generalized Principal Component Analysis: An Introduction
Speaker: Shankar Sastry
Time and Place: Tuesday 2010-04-27 13:15-14:00 in Glashuset
Abstract: There are a large number of problems in which we encounter the problem of modeling large amounts of data, by what is referred to as a "mixture of models", that is to say that the data can be segmented into finitely many sub components, each of which can be separately modeled. In the context of the identification of hybrid systems it is easy to see how this would arise when the input-output behavior depends on the "discrete state" of the hybrid system. Of course, the applications in computer vision, signal and image processing and indeed more generally in statistics are extremely numerous. This area of work has found a tremendous outpouring of effort and methods in recent years in the areas of machine learning, signal processing, hybrid systems, statistics and adaptive systems literature. However, it is our perception that the conceptual and theoretical underpinnings of the bulk of the literature could be improved.
In the course of a recent set of papers with Yi Ma of the University of Illinois, Urbana Champaign and Rene Vidal of Johns Hopkins University and their students, we have developed what we believe to be an interesting new approach to simultaneously segmenting and modeling data from mixtures of models. The heart of our approach lies in what is called "Generalized Principal Component Analysis". This in turn has many connections with such classical problems as Hilbert's Nullstellensatz and some other open problems in algebraic geometry and statistics. In my talk at this workshop, I will give a brief overview of the approaches and their applications to date. The work is being incorporated into a monograph to be published by Springer Verlag in 2011 (?).

Collaborating Swarms, Multi-network Topologies and Constrained Coalitional Games
Speaker
: John S. Baras
Time and Place
: Wednesday 2010-01-20 13:15-14:00 in Visionen

Abstract: We consider the problem of autonomous collaboration in groups of robots or vehicles or agents in general. We describe methods for deriving local coordination rules using techniques from time-varying Markov random fields, which result in distributed asynchronous coordination algorithms using parallel Gibbs samplers. The algorithms circumvent the well known problem of traditional potential methods that get stuck in locally optimal paths. We show that under reasonable and mild assumptions globally optimal coordination paths emerge from these local strategies. We then consider the tradeoffs between performance and execution time. We develop and analyze two additional distributed coordination algorithms to speed up convergence, a hybrid one which is a mixture between deterministic gradient coordination and randomized Gibbs samplers, and another one which adds memory to this second hybrid algorithm. We demonstrate that these algorithms converge much faster while still resulting in nearly optimal paths. We then investigate the role of the communication topology among the collaborating agents in improving performance of distributed algorithms on graphs, such as convergence speed. We rigorously demonstrate that Small World graphs emerge as a good tradeoff between performance and efficiency in consensus problems, where the latter serves as a prototypical coordination problem. We discuss extensions to expander graphs and the significance of separating the collaboration topology from the communication topology in collaborating swarms. Next we introduce constrained coalitional games and we show that they capture in a fundamental way the basic tradeoff of benefits vs. cost of collaboration, in networked collaborating systems. We demonstrate that various simple models of constrained coalitional games can explain network formation and the emergence or not of collaboration. We close with conclusions on autonomic networked swarms and examples from biology, engineering, social and economic networks, and provide a brief list of interesting future research directions.


23-24 November NTU-CADICS Workshop on Autonomy in Vehicular Systems, Program


3-4 November (lunch-lunch): CADICS Research retreat at Villa Fridhem



Multi-agent UAV collision avoidance, Mon 5 Oct, 15.15, Visionen

Speaker:     Dr. Michal Pechoucek, head of Agent Technology Center and Deputy head of the Department of Cybernetics, Czech Technical University in Prague
Date:        2009-10-05
Time:        15.15-16.00
Place:       Visionen, entrance B27, B-Building, Campus Valla, LiTH
Language:    English

Abstract

Free-flight concept is a modern approach to flexible collision free air traffic control investigated mainly in the context of unmanned aerial assets. Agent Technology Center (ATG) at the Czech Technical University designed and developed AGENTFLY, a multi-agent simulation environment used for a variety of coordinated UAA scenarios. Several collision avoidance methods (e.g. based on visual flight rules, based on peer-to-peer negotiation, based on multi-party interaction or prediction) and a specific A* based planning algorithm were successfully tested and demonstrated in AGENTFLY system. During my talk I will share with the audience experience collected during the AGENTFLY system development, will present several demos and will comment on industrial take up of the system.

Bio

Dr. Michal Pechoucek is a head of Agent Technology Center and Deputy head of the Department of Cybernetics, Czech Technical University in Prague (CTU). He is an Associate professor in Artificial Intelligence at CTU. He graduated the University of Edinburgh and completed his Ph.D. in Artificial Intelligence and Biocybernetics at CTU. Michal Pechoucek has acted as a principal investigator on more than 20 research projects funded by AFRL/EOARD, ONR, ARL, CERDEC and FAA. Besides, Michal Pechoucek has collaborated with international industries such as Rockwell Automation, BAE Systems, Cadence Design Systems, TSystems, Denzo and NASA. Michal Pechoucek is a visiting/honorary member of Artificial Intelligence Application Institute, University of Edinburgh, is member of the Advisory board of Center of Advanced Information Technologies at State University of New York, University of Bingahmton where he was a visiting professor, visiting postdoctoral researcher at the University of Calgary and he was a member of the AgentLinkIII European Coordinating Action management committee. Michal Pechoucek is an author or co-author of SCI cited publications in proceedings of international conferences and journal papers, PC member and a co-chair of relevant conferences and workshops. Michal Pechoucek was a chair and is a member of EUMAS advisory board and member of CEEMAS steering committee.


3-day course: Distributed and Event-based Control, September 2-4
Speakers: Anders Rantzer (LTH ) and Karl-Henrik Johansson (KTH)
Date: 2009-09-02 -- 2009-09-04
Place: Visionen and Glashuset

Abstract:
How can control synthesis for a complex distributed system be done in a
systematic way? Can a global objective for the total system be split
into individual objectives for subsystems? What information transfer is
most important for control performance? These are some of the main
issues in this course, which combines ideas for decentralized
optimization and control from the 1960-70s with recent work on networked
control systems.

Schedule:
September 2-3 (Anders Rantzer)
Wednesday
0915-1200 Lectures in Visionen
Time for individual study and problem solving
1530-1700 Exercise session in Glashuset

Thursday
0915-1200 Lectures in Glashuset
Time for individual study and problem solving
1530-1700 Exercise session in Glashuset

September 4 (Karl-Henrik Johansson)
Friday
0915-1200 Lectures in Visionen
1330-1500 Exercise session in Glashuset



Title:        The 20th International Workshop on Principles of Diagnosis
Date:        2009-06-14 - 2009-06-17

Place:    Skepparholmen conference facility, Stockholm

The International Workshop on Principles of Diagnosis is an annual event that started in 1989 rooted in the Artificial Intelligence (AI) community. Papers presented at the workshop cover a variety of theories, principles, and computational techniques for diagnosis, monitoring, testing, reconfiguration, fault-adaptive control, and repair of complex systems. Applications of these theories, principles, and techniques to industry-related disciplines and other real-world problems are also important topics of the workshop.

The complete program, including electronic version of submitted papers, is available from the homepage http://www.isy.liu.se/dx09/



Title:     Joint CADICS/MOVIII Seminar – What we don't know about Robotic Aircraft or UAS

Speaker:   Professor Rodney Walker, Queensland University of Technology and Australian Research Centre for Aerospace Automation
Date:        2009-05-26
Time:        13.15-14.00
Place:       Visionen, entrance B27, B-Building, Campus Valla, LiTH
Language: English

Abstract

The ARCAA has been conducting research in the area of robotic aircraft (UAS, UAV, ..) since 1999. In this time the group has been on journey of understanding, beginning with very little knowledge of the problems associated with these vehicle, and looking from the elec eng/computer science perspective.  This talk will address the issues of societal acceptance of UAS, followed by a discussion of current issues associated with regulatory impediments. Specific comment will be made in the Australian context. It will also highlight problems that remain to be solved for UAS.

Research results will be provided for two research thrusts; namely the emergency (or forced) landing problem and the "see and avoid" problem.  These areas being investigated for automation through the use of computer vision techniques. A brief overview and video of our flight test activities for the "Smart Skies" program (an automated aircraft separation program) and our robotic powerline asset inspection program will be provided. Finally, speculation of the future development path, and likely hurdles, for robotic aircraft will be provided.

 

Bio

Rod Walker is the Professor of Aerospace Avionics at the Queensland University of Technology and the Director of the Australian Research Centre for Aerospace Automation (ARCAA).  ARCAA is a joint venture between QUT and the CSIRO ICT Centre.  In his career he been responsible for the development of electromagnetic propagation models (parabolic equation method) to estimate the performance of the GPS in multipath significant environments.  He was also the GPS program leader on an Australia national satellite program "FedSat" working closely with the NASA Jet Propulsion Lab to deliver and operate the GPS payload.  Over the last decade he has turned his attention to the aviation environment and focuses on removing impediments to increasing automation in this context.  He is a private pilot with aerobatic endorsement.  He is on the board of AUVSI Australia and Chairs a National Committee on UAS Regulatory Development. His group has over 30 staff and a custom built facility at Brisbane International Airport.


Title:         CADICS/MOVIII Distinguished Lecture, Thursday May 28th 9.00-10.00 Wolfram Burgard

                Probabilistic Techniques for Mobile Robot Navigation

Speaker:   Wolfram Burgard, Albert-Ludwigs-Universität Freiburg, Germany

Date:        2009-05-28

Time:        09.00-10.00

Place:       Visionen, entrance B27, B-Building, Campus Valla, LiTH

Language: English

 

Abstract

In recent years, probabilistic techniques have enabled novel and innovative solutions to some of the most important problems in mobile robotics.  Major challenges in the context of probabilistic algorithms for mobile robot navigation lie in the questions of how to deal with highly complex state estimation problems and how to control the robot so that it efficiently carries out its task.  In this talk I will discuss both aspects and present an efficient probabilistic approach to solve the simultaneous mapping and localization problem for mobile robots.  I will also describe how this approach can be combined with an exploration strategy that simultaneously takes into account the uncertainty in the pose of the robot and in the map. For all algorithms I will present experimental results, which have been obtained with mobile robots in real-world environments as well as in simulation. I will conclude the presentation with a discussion of open issues and potential directions for future research.

 


 

In conjunction with Wolfram Burgard's CADICS/MOVIII Distinguished Lecture the Swedish AI Society is arranging its yearly workshop at Linköping University. In addition to a number of talks by workshop participants, the workshop features five more interesting invited speakers and a live demonstration of the humanoid robot Nao from Aldebaran Robotics.

This is a great opportunity for CADICS/MOVIII researchers to get a view of current AI research in Sweden. The registration fee is only 900SEK (students 600SEK) for the two days which includes lunch both days and a dinner on Wednesday evening. If you only want to attend stray talks then you can do so without registering for the workshop if there are empty seats. 

The complete program and registration information is available from the homepage http://www.sais.se/sais2009.

INVITED SPEAKERS:

* Wolfram Burgard, Albert-Ludwigs-Universität Freiburg

  Probabilistic techniques for mobile robot navigation

* Tom Ziemke, Skövde University

  Why robots need emotions

* Danica Kragic, KTH

  Vision for object manipulation and grasping

* Patric Jensfelt, KTH

  Spatial Modeling for Cognitive Systems

* Christian Schulte, KTH

  Constraint Programming for Real

* Magnus Merkel, Fodina Language Technology AB

   Title to be announced

Welcome!

Patrick Doherty, Fredrik Heintz and Jonas Kvarnström

 


Title:
        
Swedish AI Society Workshop
Date:        2009-05-27 - 2009-05-28

Joint MOVIII/CADICS Workshop, May 14, 2009: For posters click here.

Logga in Produced by Mediatron