Ambient Intelligence

With the emerging mass production of very small, cheap Radio Frequency Identification (RFID) tags, it is becoming feasible to deploy such tags on a large scale. In this paper, we advocate distribution schemes where passive RFID tags are deployed in vast quantities and in a highly redundant fashion over large areas or object surfaces. We show that such an approach opens up a whole spectrum of possibilities for creating novel RFID-based services and applications, including a new means of cooperation between mobile physical entities. We also discuss a number of challenges related to this approach, such as the density and structure of tag distributions, and tag typing and clustering. Finally, we outline two prototypical applications (a smart autonomous vacuum cleaner and a collaborative map-making system) and indicate future directions of research.

One proposed way to realize the AmI vision is to turn everyday objects into artifacts (by adding sensing, computation and communication abilities) and then use them as components of Ubiquitous Computing (UbiComp) applications within an Ami environment. The (re)configuration of associations among these artifacts will enable people to set up their living spaces in a way that will serve them best minimizing at the same time the required human intervention. During the development and deployment of UbiComp applications, a number of key issues arise such as semantic interoperability and service discovery. The target of this paper is to show how ontologies can be used into UbiComp systems so that to address such issues. We support our approach by presenting the ontology that we developed and integrated into a framework that supports the composition of UbiComp applications.

This paper investigates the impact of agent migration on the performance of personal agents in an ambient use case scenario. Four migration policies are proposed and a performance model, based on two quantitative performance metrics, network load and response time for processing the user request, is applied to compare the performance of the proposed migration policies. Scenario mapping to performance model and analytical results are also discussed in this paper. An agent emulator is designed and implemented in the Java programming language based on the Jatlite agent framework. The emulator is used to produce the experimental results of this study. It has been shown that, in this scenario, the policy in which a user agent follows the mobile user in the wired side of wireless network lowers the response time when the agent size is smaller than the reply size. However, when the reply size is smaller than the agent size, a simple stay-at-home policy outperforms the other three policies.

With the rapid growth of emerging demand and deployment of wireless LAN (WLAN), much of traffic including multimedia traffic is forced to travel over WLAN. Since the legacy IEEE 802.11 standard is unable to provide adequate quality of service (QoS) for multimedia traffic, IEEE 802.11e group developed medium access protocol (MAC) improvements to support QoS sensitive applications and to make more efficient use of the wireless channel. Although this QoS enhancement is great improvement for the legacy standard in terms of capability and efficiency of the protocol, it does not address the method to enhance QoS when a wireless station (STA) changes its associated access point (AP) due to its mobility. Thus, in this paper, we propose an effective QoS provision method that can guarantee QoS requirements in WLAN when the STA performs handoff and is turned on. In order to guarantee QoS requirements, the proposed method uses APs with dual radio frequency (RF) modules. By adding to AP an RF module which can only receive signals (SNIFFER), the AP can eavesdrop channels of its neighboring APs selected by the modified neighbor graph (NG). Experimental results show that the proposed mechanisms can guarantee QoS for multimedia traffic without ping-pong phenomenon by reducing the link layer (L2) handoff delay drastically and distributing the system loads.

Ambient intelligent applications require applications to recognise user activity calmly in the background, typically by instrumentation of environments. In contrast, we propose the concept of Cooperative Artefacts (CAs) to instrument single artefacts that cooperate with each other to acquire knowledge about their situation in the world. CAs do not rely on external infrastructure as they implement their architectural components, i.e. perceptual intelligence, domain knowledge and a rule-based inference engine, on embedded devices. We describe the design and implementation of the CA concept on an embedded systems platform and present a case study that demonstrates the potential of the CA approach for activity recognition. In the case study we track surface-based activity of users by augmenting a table and household goods.

We discuss the issue of privacy protection in collaborative filtering, focusing on the commonly-used memory-based approach. We show that the two main steps in collaborative filtering, being the determination of similarities and the prediction of ratings, can be performed on encrypted profiles, thereby securing the users’ private data. We list a number of variants of the similarity measures and prediction formulas described in literature, and show for each of them how they can be computed using encrypted data only. Although we consider collaborative filtering in this paper, the techniques of comparing profiles using encrypted data only is much wider applicable.

Service discovery is a process of locating, or discovering, one or more documents, that describe a particular service. Most of the current service discovery approaches perform syntactic matching, that is, they retrieve services descriptions that contain particular keywords from the user’s query. This often leads to poor discovery results, because the keywords in the query can be semantically similar but syntactically different, or syntactically similar but semantically different from the terms in a service description. Another drawback of the existing service discovery mechanisms is that the query-service matching score is calculated taking into account only the keywords from the user’s query and the terms in the service descriptions. Thus, regardless of the context of the service user and the context of the services providers, the same list of results is returned in response to a particular query. This paper presents a novel approach for service discovery that uses ontologies to capture the semantics of the user’s query, of the services and of the contextual information that is considered relevant in the matching process.

We present an access control architecture that restricts access of train and tourist office services to travelers that are really on the train. For this purpose location and velocity information are used as alternative authentication and authorization credentials. The ambient intelligence of our transparent access control mechanism manifests itself as a pool of verification methods of increasing strengths. These verification methods are based on the proximity principle and relational contextual history information about WLAN access points, fellow train travelers, the train itself and its conductor or driver.

The next generation sportsmen will be one that can be monitored and steered by a coach on their individual performances. Not only the physiological performances will be monitored, but also the contextual data, like position of sportsmen. Based on performance models, key parameters that should be monitored for rowing and soccer were identified. These parameters must be acquired real-time, processed real-time and given real-time feedback on. In this short paper, we will discuss the methods and technology developed to achieve these goals. The work presented in this short paper is still in progress.

From a User-Centered Design perspective, technology pushes are often regarded to as negative because the ideas behind these pushes not always address user needs, often causing products to fail in the market. Feasibility studies help close the gap between technology pushes and demand pulls. By inviting users to witness feasibility studies in an early stage of a design process, participants not only are able to provide input long before full functionality has been developed, but it also allows them to make that important step from imagining what an Ambient Intelligent product can do for them in their daily lives, to actually experiencing it.

Research on developing context-aware applications faces the challenge of collecting, integrating and combining information in a wide range of formats from a variety of sources, such as desktop applications, web services, user profiles and databases. On the other hand, the need for pervasive real-time application integration has boosted the development of Integration Platforms in enterprise environments. These platforms provide an infrastructure of services that enable dozens of applications spread around the organization to recognize and respond to the presence of information. In this paper we propose that integration frameworks can be used to facilitate the development of context-aware applications. As a proof of concept, we developed a simple context-aware application that works on top of a well-known integration framework.

In this article we identify the general communication patterns of physical devices and define all interfaces and conflict resolution strategies that are present in any ensemble. Based on this approach we define a generic component topology applicable for Ambient Intelligent scenarios. We validate our interface framework by showing the application of two smart scenarios: the smart conference room and the smart living room and the possibility to move applications and devices from one ensemble to the other. Our approach guarantees the fully extensibility of device ensembles with new devices.

Emerging technologies in building automation have the potential to increase the quality and cost effectiveness of services in the building industry. However, insufficient range of collected data and models of the physical and behavioural aspects of the facilities limit the capabilities of building automation systems. We describe a project for improving building services by collecting comprehensive data from variable sources and generating high-resolution models of buildings. In this context, location sensing is critical not only for data collection, but also for constructing models of buildings as dynamic environments. We first examine a range of existing location sensing technologies from the building automation perspective. We then outline the implementation of a specific location sensing system together with respective test results.

An important component of ubiquitous computing is the ability to quickly sense the dynamic environment to learn context awareness in real-time. To pervasively capture detailed information of movements, we present a decentralized algorithm for feature extraction within a wireless sensor network. By approaching this problem in a distributed manner, we are able to work within the real constraint of wireless battery power and its effects on processing and network communications. We describe a hardware platform developed for low-power ubiquitous wireless sensing and a distributed feature extraction methodology which is capable of providing more information to the user of events while reducing power consumption. We demonstrate how the collaboration between sensor nodes can provide a means of organizing large networks into information-based clusters.

To realise an Ambient Intelligence environment, it is paramount that applications can dispose of information about the context in which they operate, preferably in a very general manner. For this purpose various types of information should be assembled to form a representation of the context of the device on which aforementioned applications run. To allow interoperability in an Ambient Intelligence environment, it is necessary that the context terminology is commonly understood by all participating devices. In this paper we propose an adaptable and extensible context ontology for creating context-aware computing infrastructures, ranging from small embedded devices to high-end service platforms. The ontology has been designed to solve several key challenges in Ambient Intelligence, such as application adaptation, automatic code generation and code mobility, and generation of device specific user interfaces.

Because of their severe resource-restrictions and limited user interfaces, smart everyday objects must often rely on remote resources to realize their services. This paper shows how smart objects can obtain access to such resources by spontaneously exploiting the capabilities of nearby mobile user devices. In our concept, handhelds join a distributed data structure shared by cooperating smart objects, which makes the location where data are stored transparent for applications. Smart objects then outsource computations to handhelds and thereby gain access to their resources. As a result, this allows smart items to transfer a graphical user interface to a nearby handheld, and facilitates the collaborative processing of sensory data because of the more elaborate storage and processing capabilities of mobile user devices. We present a concrete implementation of our concepts on an embedded sensor node platform, the BTnodes, and illustrate the applicability of our approach with two example applications.

Today, most mobile devices (e.g. PDAs) are in some way associated to a fixed personal computer or server. In general this relation is only taken into account for synchronization purposes.This is rather restrictive as, while away from these fixed computers, such mobile devices may require resources that are not available (e.g. network bandwidth, processing power or storage space). This lack of resources prevents the user from doing what he wants when he wants.We propose a system in which, by enabling automatic remote code execution on a remote computer, these limitations are subdued. At run time it is decided whether some application code should run locally or on a remote computer. This is achieved using runtime meta-programming and reflection: we transform a centralized Python application so that some part of its code is run on another computer, where the needed resource is known to be available. This is accomplished without any manual code change. The performance results obtained so far, i.e. with no optimizations, are very encouraging.

In this paper we discuss research work that enables the development of mixed societies of communicating plants and artefacts. PLANTS is an EU-funded Research and Development project, which aims to investigate methods of creating “interfaces” between artefacts and plants in order to enable people to form mixed, interacting (potentially co-operating) communities. Amongst others the project aims to develop hardware and software components that should enable a seamless interaction between plants and artefacts in scenarios ranging from domestic plant care to precision agriculture. This paper deals with the approach that we follow for the development of the homonymous system and discusses its architecture with special focus on describing the communication among artefacts and plants and on designing an ontology that provides a formal definition of the domain under consideration.

In this paper we describe the issues that need to be addressed when setting up an aware environment occupied simultaneously by several users. Combining the delivery of services for various users simultaneously requires the setting up of users profiles as a record of their needs, requirements and desires. We are thus interested in a assessment of the requirements and specifications of user profiling. Furthermore there is also a need for the merger of multiple user profiles. As we are involved in the development of a smart family home and a responsive Exhibition booth, we will investigate user profiling and profile management within these two contexts. We finally discuss some issues that we consider detrimental to the success of aware environments.

In this paper we propose a tangible cube as an input device for playfully changing between different TV-channels. First we consider several design approaches and compare them. Based on a cube that has embedded gravity sensing and wireless communication capabilities a prototype is implemented. A 3D graphical representation of the cube is shown on the television screen. On each face of the cube a TV stream is rendered. The motion of the cube on the screen is connected to the rotation the user performs using the real tangible cube. Our hypotheses is that users can use the cube to browse between channels and to zap intuitively and playfully gaining a improved user experience even if the efficiency is limited compared to a remote control. We report on initial user feedback testing our hypothesis in witch we found out that users can easily use the cube without instructions and, despite technical limitations, see it as an improvement of current systems. Finally we discuss the issues that emerged from user’s feedback.

We currently witness a massive digitization of domestic materials e.g. photos, music, calendars, recipes, notes, messages. This digitization provides new conditions for how we interact with materials as well as how users shape the ambience of their homes. Observing the qualities of physical materials in the home, the process of digitization risks loosing the qualities of the spatial distribution, the aesthetics, and common reference points offered by physical materials and places in the home. We present a concept of domestic hypermedia (DoHM), which exploits potentials of digital materials and at the same time allow people to interact with digital materials in engaging ways providing rich experiences when organizing and using digital materials in homes. We present an infrastructure and design concepts that offer: ambient access to digital materials, common reference points, and collective experiences.

This paper describes an extension to Ullmer and Ishii’s TUI-categorization [41]. The reason for adding new categories is based on their omission of their associative TUIs and our work in this area of personal objects. The benefit of using personal objects instead of generic objects is that in the first instance users already have mental models or personal links between experiences, the related media and these objects. In addition, a Graspable or Tangible User Interface with personal objects can support existing media systems, instead of designing new ones that have to be learned by users.

We report research into concepts and technology for enabling end-users to configure Ambient Intelligent environments. In this paper we focus on the feasibility and acceptability of this endeavor from an end-user perspective. We describe a conceptual model and an experimental enabling technology that illustrates the viability of these concepts and a multi-faceted evaluation of these concepts from an end-user perspective. Our work suggests the need for a flexible approach in letting users choose how much should be observable of system structure and function or of the processes of system learning and adaptation. Directions for future research in this field are described in the form of some provisional principles for shaping the interaction with end-user configurable Ambient Intelligence environments.

We discuss how an augmented-reality platform can be used as a transparent interface to a windows environment. The resulting Visual Interaction Enriched Windows (VIEWs) intend to realize an evolutionary rather than a revolutionary transition from the classical desktop environment to an augmented-reality (AR) environment. Therefore, in VIEWs, windows applications can still be controlled by standard means, i.e., by using mouse and keyboard. In this way, acquired user skills with existing windows applications can still be exploited. The additional interaction styles that are offered by the AR platform, including two-handed interaction, pen input with in-place visual feedback, and transparency, may however be used to improve specific interactions, such as sketching and handwriting, that are more difficult to perform on a classical desktop. The user is free at all times to choose the interaction style that best suits his/her needs when performing specific operations. The VIEWs concept has been implemented as part of an existing AR tool for designers.

We put forth a proposal for a system that creates an augmented social environment where devices allow for an integrated capture of information about a social gathering (e.g. pictures, audio recordings, written thoughts and expression). Users can subsequently navigate around a collection of images, while listening to sounds and conversations recorded with the photo at the time of capture. We describe the devices used to capture this information and the user experience of the integrated Memory Collage application. The paper will also discuss the implications found for ambient intelligence within social environments.

The topic of this paper is rich interaction. Rich interaction borrows from tangible interaction and the concept of affordances. This is achieved through integral design of form, interaction and function of products. It is applied to interactive consumer products. A digital camera with a rich user interface (RUI) was designed and compared in a user study to a digital camera with a more conventional user interface. Several issues concerning rich interfaces are discussed.

The concept of metaphor has been used in UI design in a loose manner. There is a need to conceptually separate it from related concepts to restore the power it used to have in rhetoric. It is also important to understand the life cycle of metaphor, how it changes over time in the conceptualisation process. This is especially topical in ubiquitous computing, in which entirely new concepts and interaction styles are introduced. In this paper, we describe the use of metaphors and related concepts in theory and apply the approach in a mobile application.

This paper defines the problem space of distributed, migratable and plastic user interfaces, and presents CAMELEON-RT, a technical answer to the problem. CAMELEON-RT1 is an architecture reference model that can be used for comparing and reasoning about existing tools as well as for developing future run time infrastructures for distributed, migratable, and plastic user interfaces. We have developed an early implementation of a run time infrastructure based on the precepts of CAMELEON-RT.

Energy conservation is a critical issue in wireless sensor networks. We formulate the energy conserving routing problem as a nonlinear program, whose objective is to maximize the network lifetime until the first node battery drains out. We prove the nonlinear program can be converted to an equivalent maximum multi-commodity concurrent flow problem and develop an iterative approximation algorithm based on a revised shortest path scheme. Then we discuss the feasibility, precision and computation complexity of the algorithm through theoretic analysis, some optimization methods are also provided to reduce the algorithm running time. Performance simulation and comparison show the effectiveness of the algorithm.

Home networks and the interconnection of home appliances is a classical theme in ubiquitous computing research. Security is a recurring concern, but there is a lack of awareness of safety: preventing the computerized house from harming the inhabitants, even in a worst-case scenario where an unauthorized user gains remote control of the facilities. We address this safety issue at the middleware level by restricting the operations that can be performed on devices according to the physical location of the user initiating the request. Operations that pose a potential safety hazard can only be performed within a physical proximity that ensures safety.We use a declarative approach integrated with an IDL language to express location-based restrictions on operations. This model has been implemented in a middleware for home audio-video devices, using infrared communication and a local-area network to implement location awareness.

The purpose of AmbieSense is to provide personalised, context-sensitive information to the mobile user. It is about augmenting digital information to physical objects, rooms, and areas. The aim is to provide relevant information to the right user and situation. Digital content is distributed from the surroundings and onto your mobile phone. An ambient information environment is provided by a combination of context tag technology, a software platform to manage and deliver the information, and personal computing devices to which the information is served. This paper describes how the AmbieSense reference architecture has been defined and used in order to deliver information to the mobile citizen at the right time, place and situation. Information is provided via specialist content providers. The application area addresses the information needs of travellers and tourists.

Ambient computing as an ideal demands levels of functional attainment that have thus far not been realised. Ambient applications require that the computing application be subsumed into the everyday context in an unobtrusive manner with interaction modalities that are natural, simple and appropriate to both the individual user and their associated context. Within this paper, we consider the use of mobile intentional agents as potential key enablers in the delivery of ambient intelligent services. In particular, we compare and contrast two agent-based ambient intelligence case studies.

We investigate the application of a logical model of agency, known as the KGP model, to develop agents for ambient intelligence applications. Using a concrete scenario, we illustrate how the logical formalism employed by a KGP agent allows a person to access the surrounding ambient through the agent in a transparent manner. We evaluate our claims by implementing the resulting interactions in PROSOCS, a prototype multi-agent systems platform that allows KGP agents to be deployed as components of ambient intelligence applications.

Context-aware services platforms aim at supporting the handling of contextual information in order to provide better user-tailored services. This paper addresses our current efforts towards a configurable and extensible services platform for context-aware applications. It discusses the use of a language and ontologies to cope with configurability and extensibility aspects.

In this paper, we review different context classification systems that have been used to define elements of context. Although existing classification systems cover various types of context, in the development of context aware applications, only a few types of context have been used. In this work, we aim to build a context classification model based on Activity Theory that provides a basis both for dialogue amongst context awareness researchers and for the implementation of a context awareness architecture.

Pervasive resp. ubiquitous systems use context information to adapt appliance behavior to human needs. Even more convenience is reached if the appliance foresees the user’s desires. By means of context prediction systems get ready for future human activities and can act proactively.Predictions, however, are never 100% correct. In case of unreliable prediction results it is sometimes better to make no prediction instead of a wrong prediction. In this paper we propose three confidence estimation methods and apply them to our State Predictor Method. The confidence of a prediction is computed dynamically and predictions may only be done if the confidence exceeds a given barrier. Our evaluations are based on the Augsburg Indoor Location Tracking Benchmarks and show that the prediction accuracy with confidence estimation may rise by the factor 1.95 over the prediction method without confidence estimation. With confidence estimation a prediction accuracy is reached up to 90%.