Enabling technologies concerning hardware, networking, and sensing have inspired the development of context-aware IT services. These adapt to the situation of the user, such that service provisioning is specific to his/her corresponding needs. We have seen successful applications of context-aware services in healthcare, well-being, and smart homes. It is, however, always a question what level of trust the users can place in the fulfillment of their needs by a certain IT-service. Trust has two major variants: policy-based, where a reputed institution provides guarantees about the service, and reputation-based, where other users of the service provide insight into the level of fulfillment of user needs. Services that are accessible to a small and known set of users typically use policy-based trust only. Services that have a wide community of users can use reputation-based trust, policy-based trust, or a combination. For both types of trust, however, context awareness poses a problem. Policy-based trust works within certain boundaries, outside of which no guarantees can be given about satisfying the user needs, and context awareness can push a service out of these boundaries. For reputation-based trust, the fact that users in a certain context were adequately served, does not mean that the same would happen when the service adapts to another user’s needs. In this paper we consider the incorporation of trust into context-aware services, by proposing an ontological conceptualization for user-system trust. Analyzing service usage data for context parameters combined with the ability to fulfill user needs can help in eliciting components for the ontology.@en

Drones performing an autonomous mission need to adapt to frequent changes in their environment. In other words, they have to be context-aware. Most current context-aware systems are designed to distinguish between situations that have been pre-defined in terms of anticipated situation types and corresponding desired behavior types. This only partially benefits drone technology because many types of drone missions can be characterized by situations that are hard to predict at design time. We suggest combining context-awareness and data analytics for a better situation coverage. This could be achieved by using performance data (generated at real-time) as training data for supervised machine learning – it would allow relating situations to appropriate behaviors that a drone could follow. The conceptual ideas are presented in this position paper while validation is left for future work.@en

We observe that context-aware systems currently developed in one domain or another are mostly technology-driven, and not so much user-centric. They are often not based on a thorough analysis of the effects they produce when interacting with their context, especially regarding the contribution of these effects to user needs. We argue that a conceptual framework is needed to support such analyses. In this paper we identify the concepts necessary to define important structural aspects of a context-aware system and its context, and to formulate generalizations about effects of the interaction of the context-aware system and its context related to user needs. Using this conceptual framework, we classify context-aware systems in terms of the kinds of context assumptions that we can make at design time, and we discuss several threats to validity of a context-aware system. We believe that the proposed conceptual framework can help to better assess the utility concerning a context-aware system design. We use various examples of context-aware applications to illustrate our ideas.

@en

Drone technology can potentially be useful for land-borer security - unmanned drone missions could be performed in the sky, supported by embedded sensors and data processing. Algorithmic rules can be incorporated in the drone software to make instant decisions, whereas other decisions might be made on the ground on the basis of monitoring data received from the drone. This allows for achieving context-awareness: the operation of the drone depends on the situation at hand. The mix of algorithmic and human decision-making distributed over many components raises questions that concern accountability - who would be responsible in case of an accident or a 'wrong doing': the hardware or software developers, the ground station managers, the law (regulations) makers, or the ones who have decided to use drones in the particular situation? In the current work we analyze the usability of drones with regard to land border security, featuring benefits and corresponding accountability concerns. To achieve this, we have studied drone technology and in particular: the technical features as well as the corresponding actor-roles and relationships, considering a land-border-security-related application scenario. On that basis we have carried out an analysis from an accountability perspective.

@en