Session III.3 Dynamically Adaptive FI-Applications: Beyond Adaptive Services

ORGANISERS / CONTACTS:    

Andreas Metzger and Clarissa Marquezan (Paluno, University of Duisburg-Essen; S-Cube Network of Excellence & FInest FI PPP Use Case project)
Katarzyna Wac (University of Geneva)
David Hausheer (TU Darmstadt)

BACKGROUND

Future Internet applications will draw on the convergence of Services, Things, Contents and Networks. This means that the capabilities and features of FI applications will be provided – to a large extent – by third parties (e.g., through Internet-based software services, public sensor networks or cloud infrastructures). As a consequence, it will become of paramount importance to build FI applications in such a way that those applicationscan dynamically and autonomously respond to changes in the provisioning of services, availability of things and contents, as well as changes of network connectivity and end-user devices.Initial solutions for the dynamic adaptation of software and service systems exist. However, those solutions need to be significantly augmented, improved and integrated with a complete system perspective. Specifically, due to the very large scale of FI applications, this requires significant progress towards distributed and highly dispersed adaptation strategies and solutions.

OBJECTIVES / DESCRIPTION

This session will elaborate on the need for such adaptation capabilities by introducing and scrutinizing representative, cross-cutting application scenarios from various (possibly complementary) domains.

The results of this session will be a collection of the presented material and a consolidated list (possibly in the form of a mind map) of research challenges towards dynamically adaptive Future Internet applications. The results will be made available on the web portal of S-Cube at www.s-cube-network.eu.

Envisioned application scenarios: We currently envision at least the following two application scenarios (involving
key elements of the Future Internet):

eHealth: The paradox of health and life expectancy in the 21st century is that while advancements in technology and medicine enable us to live longer, our modern lifestyle habits increase the probability of getting chronically ill and experience long-term limitations, until death. Yet, current legacy health systems are designed more for acute cure rather than continuous chronic care, which in turn is leading to a continuous increase in health care costs. An integrated health care approach supported by Future Internet technologies anytime-anywhere-anyhow is a key to achieve the ambitious goal of efficient and effective managing of diseases and enabling the self-management of patients ‘on-the-move’. This poses requirements on the Future Internet in terms of secure distributed data management and privacy of the patient’s location and health state, as well as a high responsiveness of the underlying network infrastructure, bandwidth on demand capabilities, and viable business models for Quality of Service assurance of eHealth applications. A key issue is the adaptation of applications to the underlying network conditions by selecting the best available quality for the particular needs of an application or by changing the application behavior.

Transport and Logistics: Future Internet technologies can facilitate radical improvements not only for optimizing existing processes in international transport and logistics, but also for radically new and innovative logistics processes. Specifically, the operational procedures of the involved parties can be significantly improved. This means that intermediate steps currently needed due to the lack of control and visibility on the logistics networks can be rationalized, which in turn leads to significant economic and social-ecological benefits. Enabling those innovations requires managing highly complex information and operational processes along with managing multiple interdependencies. Several technological innovations, such as RFID or EPC global are a first step towards such optimized transport and logistics processes. However, to become truly efficient, a key issue will be predictive, distributed event handling capabilities (exploiting information from sensor networks and video streams) that allow for the treatment of delays and other unforeseen events by proactively adapting the logistics processes.

Media: Most media distributors have migrated from analogue to digital media formats, but their applications are still static and inflexible and based on proprietary applications and devices and working with files and streams. They can greatly increase business flexibility to address new opportunities using the Future Internet by encapsulating content as media objects that can be composed and manipulated using services and can actively support the de-livery of content from its creators to consumers. Media producers understand what makes content attractive to consumers, and how to produce it while meeting business constraints on quality and cost. But they don’t understand how to compose services, or how to use composed services to manage content delivery in an extended supply chain connecting content creators to consumers. Imagine that a media producer wants to deliver a high-definition movie to consumers who will view it using PC or smart phone devices. To make money from the service, advertising agencies will be invited to insert targeted content at broadcast time, bidding for slots based on the number of consumers, and collecting their responses through the distributor’s social networking site. The producer describes what she wants to a service engineer from her Internet division, who sets up the system in collaboration with counterparts from the most important advertisers. The programme goes out at the appointed time and everything seems to go well. But unfortunately a Vodka product placement causes grave offence to Muslims in Saudi Arabia, and to parents in the USA where kids saw it as they came home from school. Later it is realised that the social networking site became overloaded by all the complaints. Many viewers in Europe were unable to connect to the site, so the advertising agencies didn’t place any targeted advertising content. The system couldn’t adapt to support mass marketing advertising content from different agencies instead, so ad revenues were well below expectations. And mobile consumers could only get a very low-resolution stream, so those with high-end mobile devices were disappointed and probably won’t tune in again.

FORMAT OF THE SESSION

This session consists of two major parts. This session starts with a short introduction and motivation for dynamic adaptive services in Future Internet applications and the presentation of application scenarios. The second part of this session is an open panel discussion involving invited panellists and Q&A from the audience, followed by the concluding remarks and follow up actions coordinated by the organizers.

AGENDA