You are here: Resources > FIDIS Deliverables > Profiling > D7.7: RFID, Profiling, and AmI > 

D7.7: RFID, Profiling, and AmI

Introduction  Title:
 Generic Understanding of AmI-Systems from a Technical Perspective


RFID, the ‘Internet of Things’ and autonomic profiling



From March to September 2006 the European Commission has initiated a Public Consultation on RFID and ‘The Internet of Things’, having invited a number of stakeholders to exchange information on the technological state of the art and to share their viewpoints on the wider implications of a tagged environment. During this consultation it became clear that RFID systems* form a potent enabling technology for AmI or networked adaptive environments. Far beyond the state of the art of supply chain management, RFID-applications were for instance presented as smart solutions for critical infrastructures, e.g. flood alert systems. Crucial infrastructures will profit from the fact that RFID systems* can provide an X-ray vision of presently invisible (undetectable) layers of reality. This is the case because the borders between the online and the offline world begin to blur as THINGS go online, creating what is now called The Internet of Things (ITU, 2005), or, as Adam Greenfield (2006) says, creating the dawn of ‘everyware’. As it was repeated during many of the presentations, the ‘Internet of Things makes devices smart, which are not necessarily smart by themselves, but smart because they are connected. These connections allow machine-to-machine (M2M) communication, leading to an environment that functions like a unified interface, producing a very different case than explicit control. M2M communication in fact produces autonomic computing, implicating real-time monitoring and real-time decision-making without human intervention. Evidently, such autonomous processes challenge our sense of privacy, control and personal autonomy. In the next section a brief analysis will be made of autonomic computing and the autonomic profiling it implies.

Autonomic computing and autonomic profiling

In 2001 Paul Horn, vice-president of IBM, coined the term autonomic computing. His concept basically refers to the process of interconnected processing of data, gathered from ‘everyware’, involving continuous M2M communication and M2M decision-making. What is special about the concept of autonomic computing is the focus on the self management of the network (Kephart, Chess, 2003). The intelligence that evolves from such a network implies pro-active instead of inter-active computing: We are not asked to program our preferences, the whole idea is that we need not interfere because the ‘everyware’ infers our preferences even before we become aware of them. Paul Horn’s choice of the term ‘autonomic’ was inspired by the subconscious functioning of our autonomous nervous system that is likewise pro-active and manages both our continuous adaptation to the internal and external environment and its own repairs. Autonomous computing seems to represent a shift from human to digital butlers, always unobtrusively anticipating one’s need one step ahead. However, to enable this butlerisation of the material environment we need a permanent data shadow, infinitely sharper than human memory, never fading, presenting us with an effective denial of oblivion.

Autonomic computing presumes autonomic profiling, which could be defined as a reiterative process of construction and application of profiles, entangling real time monitoring and real time M2M decision making. The adaptive environments envisioned in AmI depend on such autonomic profiling, restricting human intervention to revisions of the software. Such revision seems marginal, because autonomous computing depends on auto-revision of the software by the network itself. It may be the case that human intervention will eventually focus on the end-user who wants to introduce deliberate changes in his relationship with the adaptive environment. Most probably the end-user will need autonomic devices to reset his default positions. This could mean that the time will come that only those humans that have access to the right type of autonomic devices can interact with their environment in an autonomous way.  


Privacy and autonomy in the age of ‘everyware’

If the reader is not familiar with concepts like autonomic computing, ‘everyware’, etc. he may feel nauseated by the fantastic perspectives presented in this section and even take a sceptical stand against such apparent ‘belief’ in the powers of technology. The point is, however, that our environment may be changing both swiftly and radically (Garreau, 2004). If these technologies take on, they may create a technological infrastructure that will impact our lives in existential ways. This is not to confess to technological determinism, but to acknowledge that technologies such as RFID systems* may come to determine fundamental aspects of our life. In section 5.3 the difference between technological determinism and different strands of constructivism will be further discussed. The awareness of the widespread social implications of a technology once it has taken its place in society is the reason why many of the presentations at the Public Consultation referred to privacy by design: We should NOT wait until the infrastructure is a fait accompli, but weave the technical possibilities for adequate end-user controls into the technologies as they emerge and enter the market. As many speakers remarked, data protection legislations seem hopelessly impotent as long as the technological means to implement its principles are absent. Some speakers argued for a mandatory privacy assessment of new technologies and mandatory introduction of PET* applications. At the same time the focus on individual deliberate consent, purpose limitation seems inadequate in the face of a technological infrastructure that aims to relieve us from the burden of deliberate interaction (replacing this with invisible pro-active M2M communication) and builds on total correlatability of ‘everyware’ (including all our data all the time everywhere). The focus on personal data instead of on electronic footprints seems to render data protection legislation outdated and ineffective as to the real threats to be faced. Both the legal and the technological infrastructure need crucial updates to recreate the adequate safeguards for constitutional democracy. In this deliverable the focus will be on a multidisciplinary description of the technological state of the art, use cases and prospective analysis, followed by an analysis of the existing legal framework and a survey of social aspects. In the new Work Package 12 on emerging technologies, especially in deliverable 12.3 a holistic framework will be elaborated to meet some of the challenges relating to RFID. In deliverables 7.8 and 7.9 the challenges of autonomic profiling will be studied in terms of the need for a closer integration of legal and technological frameworks.



Introduction  fidis-wp7-del7.7.RFID_Profiling_AMI_02.sxw  Generic Understanding of AmI-Systems from a Technical Perspective
Denis Royer 4 / 43