Participants should arrive on Sunday (August 15, 2010) between 3 p.m. and 7 p.m., and leave on Saturday (August 21, 2010) after lunch at around 2 p.m. The official program will start on Monday (August 16, 2010) 9 a.m. and end on Saturday (August 21, 2010) 12.15 p.m.
Sunday, August 15
15.00-19.00 Arrival of participants and light dinner
Monday, August 16
08.45-09.00 Kay Roemer, Friedemann Mattern: Welcome (slides)
15.30-16.00 Coffee & Cake
19.30 Art exhibition Visual Japanese Poetry opens - all participants of the summer school are invited to join
Tuesday, August 17
09.00-10.30 Friedemann Mattern and Thorsten Staake: Smart Energy I (slides)
15.30-16.00 Coffee & Cake
Wednesday, August 18
09.00-10.30 Participants Workshop I
- Antonio Guerrieri: A WSN-based Building Management Framework to Support Energy-Saving Applications in Buildings
- Markus Weiss: Creating user awareness: Changing energy usage with the help of a smart energy system
- Richard Mietz: Searching the Future Internet
- Delphine Christin: Privacy in Participatory Sensing
- Sonia Waharte: Search and Rescue with Unmanned Arial Vehicles
- Yasir Javed: RTRACKP: Distributed Multi Robot Tracking System Using Wireless Sensor Networks
10.45-12.15 Participants Workshop II
- Amro Al-Akkad: UbiComp for First Response: Context-Aware Mobile Sensing for Restoring Security and Safety in Crisis
- Raffaele Gravina: Domain-specific approach to support WBSNs programming: the SPINE Framework
- Peter Kreyssig: Chemical Computing Methods for Wireless Sensor Networks
- Pablo Guerrero: Routing Metrics for In-Network Workflow Execution
- Sylwia Romaszko: QoS Ad Hoc Routing Protocol Analysis in Civil Safety Context
- Michael Walsh: On the effect of communication constraints on robust performance for a practical 802.15.4 WSN
- Boattrip on river Saar
- Short hike to the Saar Loop (about 1h, bring sturdy shoes)
- Wine tasting and dinner at Schloss Ziegelberg
Thursday, August 19
15.30-16.00 Coffee & Cake
Friday, August 20
15.30-16.00 Coffee & Cake
Saturday, August 21
13.30 Departure of participants
Lectures & Hands-on Courses
Prof. Alejandro Buchmann, TU Darmstadt
Complex event detection and notification in mobile sensor-actor networks (lecture)
The behavior of sensor-actor networks is well described by Boyd’s OODA cycle. The Observe-Orient-Decide-Act cycle was developed by Boyd, a former fighter pilot, to describe the behavior of reactive systems. In the Observation phase primitive events are detected; in the Orientation phase they are combined, enriched with context information, and higher level events are derived; in the Decision phase the best course of action is determined based on the results of the Orientation phase; and in the Action phase decisions are played out yielding new events and situations.
The critical phase is the Orientation phase where complex events are derived and contextualized. This phase is particularly difficult in a distributed environment with mobility, potentially unstable communication and arbitrary delays. In this lecture we will cover the operators of event algebras, the semantics of different event consumption modes, mechanisms for combining infinite streams of events with finite relations, as well as, problems associated with synchronization, false positives and false negatives, event lifecycle, event dissemination and security.
Prof. Peter Corke, Queensland Technical University
Large-Scale Outdoor Sensing and Actuation Networks (lecture)
The talk will discuss recent work in creating large-scale sensing and actuation networks for novel applications. The first part of the talk is about large-scale environmental monitoring using traditional wireless sensor networks that measure scalar variables such as temperature and humidity. Audio or video sensing and in-situ processing presents real challenges for mote-class hardware and necessitates architectural extensions for the node. Mobility is an important aspect of the total sensing problem and can be classified as either natural and engineered mobility. One example of natural mobility is large-scale monitoring of cattle wearing sensor network devices which allows real-time monitoring of their individual and social behaviour patterns. Engineered mobility is achieved by linking sensed information to motion control and two examples will be considered: actuation of cattle to stay within defined spatial bounds, and a robotic watercraft that interacts with a large-scale fixed network of floating sensor nodes.
Prof. Hartmut Hillmer, University of Kassel
Micromachining: Foundations, Technological Fabrication and Selected Applications such as "a Sunglas for the Office Window" (lecture and experiments)
The goals of this lecture with experiments are to learn (i) basic principles of micromachining technology including specific processes and methodology. (ii) application potential of micro-electro-mechanical systems (MEMS) and optical MEMS, (iii) technology transfer in a specific example: the intelligent window. Content:
- Introduction to modern fabrication processes including sacrificial layer technology and intelligent structure release by e.g. critical point drying.
- Introduction to micromachining and microsystem technologies. Why miniaturization? Supported by some experiments to demonstrate that increasing miniaturization provides growing mechanical stability, longer lifetimes, less material fatigue and more efficient actuation
- Sensors and actuators, working principles and fabrication technology
- Examples: from simple membranes, springs, resonator elements, cantilevers, valves, manipulation elements and gripping tools finally to light modulators, optical switches, projection displays, micro optical bench, micromachined tunable lasers and filters
- The intelligent window: Operation principle, Fabrication process and Applications (Light steering in buildings, energy saving, light concentration on photovoltaic modules, smart personal environments, medical care, safety, display function).
Dr. Stamatis Karnouskos, SAP Research
Mobility and Cooperation in Real-World Enterprise Services (lecture)
Next generation enterprise services will highly depend on timely information acquired by the real world, that can be analyzed and provide business insights. This implies cross-layer cooperation among services and largely heterogeneous systems of networked embedded systems, even at device level. Several promising efforts are already underway in several domains, one of which is the future Energy domain i.e. an infrastructure for the smartgrid, the smarthouse, the electric car etc. This is a highly heterogeneous system where mobility will play a key role, and where networked embedded devices and services depending on them are expected to lie in its heart. In order to realize the promise, a key element would be to have timely monitoring and control. These coupled with open information exchange will boost cooperation among the different entities, as they will be able to access and correlate information that up to now either was only available in a limited fashion (and thus unusable in large scale) or extremely costly to integrate. This has profound impact on existing industry processes, business models and offered market services.
Prof. Akos Ledeczi, Vanderbilt University
Acoustic and RF Localization: an Overview (lecture)
There is a wide variety of approaches to localization in wireless sensor networks. They all have different properties, advantages and shortcomings. This is still a very active research area since we have yet to see a universal solution that meets the requirements of all -or even most- WSN applications. In this lecture, we will survey the state-of-the-art of acoustic and RF-based methods that have been the most widely studied to date.
Prof. Erik Maehle, University of Luebeck
Towards Underwater Robot Swarms (lecture)
Underwater Robots can be roughly divided into Remotely Operated Vehicles (ROVs) and Autonomous Underwater Vehicles (AUVs). Typical applications are inspection of pipelines or wrecks, environmental monitoring, underwater construction or repair, underwater research (in particular in the deep sea) or searching for sea mines. AUVs have to face a lot of challenges not present for other mobile robots. They have to operate in 3D space while facing high pressure and hydrodynamic forces which are both nonlinear and unpredictable. Many advanced sensors like infrared laser scanners cannot be used, also cameras have only a very short range depending heavily on the light and water conditions. Electromagnetic communication is not possible except at very short ranges, in particular there is no GPS for self-localization. Mainly acoustic communication and positioning systems are available which are, however, characterized by low bandwidth and high error rates. This is one of the main reasons that only very little work has been done so far on cooperating underwater robots or even underwater robot swarms.
In the talk the AUV Monsun currently developed by our institute is introduced in more detail. Monsun is rather small (about 40 cm long) and has been designed for low-cost AUV swarms for environmental monitoring tasks in shallow waters like the Baltic Sea. A small and low-cost underwater modem has been developed for reliable robot-to-robot or robot-to-buoy communication. Typical scenarios for self-localization and cooperation among robots or with stationary nodes for such underwater robot swarms will be presented and discussed.
Prof. Erik Maehle and Raphael Maass, University of Luebeck
Cooperation of Mobile Robots and Sensor Nodes (hands-on tutorial)
The goal of this hands-on tutorial is it to offer an introduction to the control and steering of mobile robots in cooperation with sensor nodes. A custom made mobile robot platform with two wheels, 15 distance sensors, a camera and an attached Tmote sky serves as basis for this tutorial. The robotic platform is controlled by a netbook running Windows XP. Throughout the tutorial C# is used as programming language. During the sample scenario the mobile robot operates in an environment, where sensor nodes mark points of interest. Once a node indicates an activity at such a point, the robot has to move towards the node and check the surrounding.
Prof. Alcherio Martinoli, EPFL
Distributed Sensing using Multi-Robot Systems: From Coverage to Search (lecture)
Technological advances in communication, embedded computing, energy storage, sensors and actuators enable an increasingly ubiquitous deployment of distributed, mobile, robotic systems in the real world. Several of the potential applications for multi-robot systems are concerned with distributed sensing and can be casted as a search or coverage problem. In this lecture, I will summarize model-based and data-driven methods that have been developed to tackle these two partially overlapping classes of problems. I will support the discussion with specific cases studies characterized by different constraints (e.g., volume, hardware and software resources) and sensing modalities (e.g., olfaction, hearing, vision) of the single robotic node, various team sizes, and different mission details (e.g., performance metrics, environmental peculiarities). I will in particular emphasize how often several extremely interesting search and coverage algorithms achieve substantially deteriorated performances when deployed on real, noisy, resource-limited, distributed platforms. Finally, I will conclude my lecture with the description of what I believe will be some of the future grand challenges in design and optimization of multi-robot systems for distributed sensing applications.
Information and communication technology (ICT) consumes energy, but is also an important means of conserving energy. In fact, it is expected that ICT will play a critical role in supporting the upcoming paradigm shifts within the energy sector towards more sustainable generation and use of electricity.
However, beyond improving the ways energy is generated, distributed, and used, ICT can also play a vital role in supporting and motivating consumers to make better decisions when pursuing energy intense activities and thus help to diminish the carbon footprint of these activities. Pivotal for that are "smart" technologies (such as wireless sensors or Web-ready devices) that can help to reduce the growing levels of energy consumption, for example in the domestic sector.
The lecture will provide some background information on global energy supply and demand, elaborate on the role ICT plays as a consumer of electricity, describe how a smart grid and smart metering can change electricity supply and usage, and outline how to combine insights into ICT and consumer research to motivate and help people leading a sustainable life.
Prof. Anibal Ollero, University of Seville
Cooperating Aerial Robots and Wireless Sensors and Actuator networks (lecture)
This lecture will present concepts, technologies and applications on the cooperation of robotic systems, and particularly aerial robots, with wireless sensor and actuator networks.
On the one hand the above cooperation includes the consideration of the robots as mobile nodes of the wireless networks and on the other hand, the wireless sensors and actuators in the infrastructure can be considered as extensions of the sensors and actuation capabilities of the robots.
The lecture will include methods to compute optimal trajectories of the robots to collect data from the static nodes and the implementation by using both aerial (fixed wing autonomous aerial vehicles) and ground robots. The cooperative perception by using aerial robots (autonomous helicopters) and wireless sensor networks, and its application to detection and monitoring of events will be also analyzed.
Moreover, the talk will include the cooperative actuation, as for example in the joint transportation of a load by means of the coordinated control of several autonomous helicopters, and the cooperation of ground actuators, such as fire extinguishers, in fire fighting activities.
The lecture will show videos and results of European and national projects with real demonstrations in several application fields such as disaster management, natural environment conservation and filming.
Prof. John Stankovic, University of Virginia
Robustness, Openness and Real-Time: Cyber Physical Systems Issues and Solutions (lecture)
Many exciting and next generation Cyber Physical Systems (CPS) will be based on wireless sensor networks as an underlying infrastructure. Many CPSs such as home health care and home security will operate in open environments and co-exist. The interactions in and across these systems will further enable new and important applications. While deploying CPSs in open and uncontrolled environments provides many benefits, it also gives rise to greater complexity, uncertainty, non-determinism, and privacy and security issues than found in today’s embedded systems. New adaptive and robust solutions that operate in real-time must be developed to handle these complexities. Examples of the difficult CPS issues in wireless communications, sensing, and control regarding robustness, real-time, and openness will be presented. First steps at solutions and solution directions will also be discussed for various topics including localization, real-time scheduling and control, run time assurance, programming abstractions, and privacy.
Dr. Michael Rohs, T-Labs / TU Berlin
Developing Mobile Sensing Applications with Google Android (hands-on tutorial)
This tutorial introduces Android as a modern platform for developing mobile applications. We will cover key aspects of Android's architectural components that are used to structure applications and communicate with the system. Based on this, the tutorial will briefly introduce user interface development in Android. It will then focus on sensor input, including how to use the integrated sensors and how to communicate to external sensors using Bluetooth. We will also introduce the Arduino platform for prototyping sensor-based applications.
Prof. Kay Roemer, Uni Luebeck / ETH Zurich
Time Synchronization for Networked Embedded Systems (lecture)
Time synchronization is a fundamental service in sensor networks, for example, to fuse sensor data from different sensor nodes, or to coordinate sensor nodes for access to the communication channel. In this lecture we discuss fundamental challenges and approaches for time synchronization in sensor networks. We also present and discuss concrete synchronization algorithms used in sensor networks.
Prof. Eduardo Tovar, Polytechnic Institute of Porto
Densely Instrumented Physical Infrastructures (lecture)
Large-scale and dense sensor/actuator deployments pose fundamental challenges concerning both interconnectivity and processing of huge quantities of information. Think about the simple example of obtaining the minimum value among cents of sensor readings. Or think about the more sophisticated (and futuristic) active flow control application where through proper modulation of aircraft skin surfaces a significant reduction of drag and related fuel consumption (and emissions) may be attained. Currently available approaches for data processing in such large-scale very dense deployments of sensors lead to energy-waste and long response-times from sensing to actuation. This lecture will address emerging techniques that are able to allow scalable and efficient data processing in large-scale dense cyber-physical systems, where cents of nodes may coexist within the same broadcast domain.
Using wireless sensor networks (WSNs) for auditing and managing the energy consumption in a building is an emerging research area. This talk defines the specific requirements for applications of energy management in the building context and proposes a novel framework for building management (BMF) to support heterogeneous platforms.
Consumption transparency and timely feedback are essential to achieve energy savings in a residential environment. We describe an easy to use interactive system that provides such consumption feedback for typical household appliances. To achieve this, we built on the capabilities of a smart electricity meter which we extended with a Web-based WLAN interface to make it a first-class citizen of a "Web of Things". We also developed a mobile phone client that allows users to connect to the smart meter and to monitor and measure the consumption of their appliances on an engaging graphical interface. With its low usage barrier and its real-time feedback on appliance level, our system allows for quickly identifying the biggest energy guzzlers and helps users decrease their energy consumption.
Richard Mietz: Searching the Future Internet
The emerging Future Internet will bring a number of new challenges due to the integration of a plethora of distributed heterogenous devices such as sensor networks, smartphones, and other sensor equipped embedded systems. The project Real-World G-Lab will contribute to solutions of these challenges by developing algorithms and techniques from low level energy efficient protocols to high level application development. My work is focussed on real-time search for the real-world states derived from sensors and other devices.
Delphine Christin: Privacy in Participatory Sensing
The sensing process in Wireless Sensor Networks is no more limited to only dedicated sensor platforms like TelosB, but includes also sensors such as accelerometers or microphones embedded in personal end devices. Designed as "Participatory Sensing", the utilization of millions of smartphones and their embedded sensors open the door to innovative applications that remain unfeasible with only dedicated hardware platforms. However, the information collected by the personal end devices raises privacy issues. In particular, the private sphere of device owners may be affected during the sensing process, and therefore has to be protected in order to encourage their contribution to participatory sensing projects. Within the scope of this talk, existing application scenarios are presented and some threats to privacy are highlighted. Our approach and work in progress are then briefly addressed.
Sonia Waharte: Search and Rescue with Unmanned Arial Vehicles
Tracking a moving target and reporting its state is the focus of several sensor network applications, ranging from search-and-rescue to fleet tracking, animal monitoring, security and surveillance. Most existing research relies on the assumption that a fixed sensor network is in place, and focuses on energy-efficient algorithms for activating sensors and relaying their data back to a fixed ground station. A fixed sensor network, however, is often costly to deploy and sustain. Moreover, sometimes there may be an urgent need to track a target in areas where no sensor network is available. The intent of this talk is to share the experience we have acquired with the deployments of UAVs for search and rescue operations. I will present the research challenges we had to address, describe the platforms we used, the experiments we conducted, and the different obstacles we encountered in the test phases. I will finally show the results we have obtained so far.
Yasir Javed: RTRACKP: Distributed Multi Robot Tracking System Using Wireless Sensor Networks
The integration of mobile robots with wireless sensor networks has emerged as a research paradigm destined to create numerous new applications. The inherent problem of embedding intelligence in a multi-robot system with high mobility can be solved by embedding the intelligent in the environment using wireless sensor networks, thus, enabling robots to concentrate only on data analysis and mobility. The applications of such integration are envisioned in realization of smart buildings, environments, (i.e. intrusion detection, target tracking, etc. R-Track project focuses on integrating robots and wireless sensor networks for intelligent monitoring and navigation. The main objectives of our work are: a) design a system to provide real-time tracking of multiple cooperative robots using wireless sensor networks, b) provide localization accuracy, ensuring real-time guarantees, c) reducing the ratio of false alarms, 4) design and development of distributed processing paradigm in the wireless sensor network to assist the cooperative work of the robots. In this talk we shall introduce R-Track project and shall present our work-in-progress. We shall also share our accomplishments so far, as well as the challenges and real life development and deployment difficulties faced during the process.
Amro Al-Akkad: UbiComp for First Response: Context-Aware Mobile Sensing for Restoring Security and Safety in Crisis
With the proliferation and trials of more and more small, smart and mobile devices Ubiquitous Computing (UbiComp) has faded into being deployed for domains that make high demands. In light of this UbiComp has the potential to foster crisis response and management: mobile handheld devices can collect over ad-hoc networks context-Aware sensor data. After gatering this data it can be shared, analysed, and visualized for supporting to restore security and safety during crisis response and manaegement. This talk will sketch challenges, ideas, methodologies of how to apply UbiComp on different end-users (first responders, ...victims, bystanders) inside disaster sites.
Raffaele Gravina: Domain-specific approach to support WBSNs programming: the SPINE Framework
Wireless Body Sensor Networks (WBSNs) enable a broad range of applications for continuous and real-time health monitoring and medical assistance. Programming WBSN applications is a complex task especially due to the limitation of resources of typical hardware platforms and to the lack of suitable software abstractions. To address this issue we have proposed SPINE (Signal Processing In-Node Environment), a domain-specific framework for rapid prototyping of WBSN applications, which is lightweight and flexible enough to be easily customized to fit particular application-specific needs.
Peter Kreyssig: Chemical Computing Methods for Wireless Sensor Networks
Artificial chemistries are used to develop a programming model which supports typical properties of biological systems like scalability, fault tolerance and self-X properties. This can be achieved by deriving global properties of the chemistry from local rules via so-called chemical organisation theory. One idea is to combine this with the concept of fraglets, a string rewriting system inspired by chemistry created by Christian Tschudin, in order to generate new algorithms and protocols and to have a more convenient basis for implementations. The talk gives a short introduction to both concepts as well as ideas for further research.
Pablo Guerrero: Routing Metrics for In-Network Workflow Execution
Our primary goal is to investigate the distributed execution of a workflow in an autonomous, self-configuring sensor/actuator network. This environment differs signifficantly from those traditionally discussed, where either source-to-sink or their reverse paths are the most relevant (Directed Diffusion, PSFQ, LEACH). In contrast, the autonomous workflow execution is carried out by one or more dynamically chosen nodes, who are notified about detected events and request the execution of actions. We inspect different metrics which are suitable for the emergent bidirectional traffic, concretely RSSI, LQI, and the more recently adopted ETX, and compare them to the traditionally used minimum hop count metric. We evaluate them in the context of the construction of a minimum spanning tree, through which events and actions flow from sensors, through workflow managers, to actuators.
Sylwia Romaszko: QoS Ad Hoc Routing Protocol Analysis in Civil Safety Context
An investigation of quality of service (QoS) approaches supporting a routing protocol in civil safety context is shown in this talk. This work is performed within the RISC (Réseaux hétérogènes Intelligents pour Situations de Risques) project that focuses on heterogeneous networks, where mobile and static nodes have the ability to monitor the environment and some so-called sink nodes forward data outside the network. Such networks could be used in a safety operation context e.g. a fireman uses radio systems and sensors to monitor its human biological constant by remote systems allowing the leader to be informed of his health or firemen could put sensors for environmental monitoring like temperature changes, presence of toxic gases or even detecting life in a building fully covered by smoke.
Michael Walsh: On the effect of communication constraints on robust performance for a practical 802.15.4 WSN
The provision of reliable communication in resource constrained Wireless Sensor Networks is a challenging problem. In this work the inherent tradeoffs and compromises that arise in this problem space are addressed using Anti-Windup (AW) techniques and validated using a benchmark IEEE 802.15.4 WSN. The joint problem of guaranteeing floor levels on Quality of Service (QoS) while simultaneously minimising power consumption is addressed in a system theoretic context through a novel extension of recent results from the AW literature. By limiting how often data is transmitted in the network or decreasing the transmit power for each node accordingly, quantifiable improvements can be demonstrated in terms of reduced overall outage probability and power consumption. Nonlinear, naturally occurring, hardware and communication constraints are explicitly considered in the design methodology. It is highlighted, using empirical evidence, how AW can limit performance degradation in the face of these potentially destabilising nonlinearities, while providing for specified levels of energy efficiency and QoS. A distributed robust power control technique is presented that employs AW techniques in a closed loop structure to address transceiver output power hardware saturation and quantisation constraints. Quantitative Feedback Theory is employed in the linear stage of the design process that demonstrates how pre-specified levels of robust performance and stability can be guaranteed in a WSN setting.