FIRE/STREP Project HOBNET
(HOlistic Platform Design for Smart Buildings
of the Future InterNET - www.hobnet-project.eu)
“Challenges and Methodologies
Towards Federated EU-Japan IoT Test-beds”
Prof. Sotiris Nikoletseas
U. of Patras and CTI
Greece
(EU-Japan Workshop, Brussels, April 18, 2013)
Overview
A. WSN Test-beds and the HOBNET Project
B. IoT Test-beds: Main Challenges
- standardization
- interoperability
- security/trust
C. Potential Methodological Approaches
- architectural designs
- cloudification
- virtualization
- crowdsourcing
D. Potential themes for EU-Japan Cooperation
A. HOBNET Main Objectives
a) an all IPv6/6LoWPAN infrastructure of buildings and how IPv6
can integrate heterogeneous technology (sensors, actuators,
mobile devices etc)
b) 6lowApp standardization towards a new embedded application
protocol for building automation
c) novel algorithmic models and scalable solutions for energy
efficiency and radiation-awareness, data dissemination,
localization and mobility
d) rapid development and integration of building management
applications, and their deployment and monitoring on FIRE test
beds
HOBNET Partners/Approach
- 4 academic groups (U. of Patras/CTI, U. of Geneva,
U. Edinburgh, U. College Dublin)
- 2 industries (Ericsson, Sensinode)
- 1 end-user (Mandat International)
- Methodological Approach: We take a holistic
approach addressing critical aspects at
different layers (networks, algorithms,
applications/tools) in an integrated way.
Implemented smart/green scenarios
•
•
•
•
•
•
•
•
•
•
•
•
Local adaptation to presence
Emergency management
Electric device monitoring
CO2 monitoring
Maintenance control
Customization
Building 3D visualization & monitoring
Mobile phone ID
User awareness
Oil tank monitoring
Garden watering
Resources tracking and monitoring
The MI HOBNET test-bed
The UNIGE HOBNET test-bed
The CTI HOBNET test-bed
Main concrete results/Exploitation
• 35% reduction of energy consumption
• Ability to select the energy saving/comfort trade-off
• Exploitation:
- rich standardization activities (IETF, ETSI M2M and One
M2M)
- deployments in highschools
- major strawberry plantation (smart watering)
- major brewery factory (Heineken group)
- a spin-off created (OptSense)
B. Challenges for IoT Testbeds
A. Standardization
• Revisiting fundamental issues in Low Power & Lossy
Networks e.g. IPv4 -> 6LoWPAN/IPv6, HTTP-> CoAP, etc
B. Interoperability
• IoT requires that they seamlessly and directly
communicate with each other and the Internet (e.g.
M2M communication)
C. Trust (not just Security)
• Especially towards active end users involvement
• Value of personal data, anonymity, privacy, identity
management, open data, reputation mechanisms
Challenges for IoT Testbeds (II)
D. Mobile test-beds, easy of deployment, “plug and
play” nature
• To exploit FIRE test-beds outside academic
environments
E. Multidisciplinarity
• Economists (market analysis, business models,
incentives mechanisms, )
• Sociologists (analyze driver and barriers to
technology adoption, models for societal value
creation)
C. Potential Methods - Architectures
• RESTful Architectural Style – Compatibility, seamless
interconnection with the Internet
• Embedded systems (e.g. WSN) are abstracted as webresources (Constrained Application Protocol, easy to
proxy from/to HTTP, every resource is identified by a URI)
+ 6LoWPAN (IPv6 over Low-Power Wireless Area
Networks)
• Embedded functionalities are represented as web
services
A HOBNET Example
• BMS for smart/green buildings
• Sensors and actuators represented as resources in
Resource Directory
• External (non-technical) users may compose their
custom use-case scenarios by combining resources in
logical expressions
Methods -Virtualization
• Virtual layers enable bi-directional interactions from IoT
nodes to applications and vice-versa
• Virtual layers are used to expose functional aspects and
information on IoT nodes as services
• They allow to organize diverse sub-networks in a
homogeneous way
A Suggested Approach
• Organize several IoT networks under a virtual network
• End users are offered a unique interface of interaction
• A meta-layer provides access via an open interface,
regardless of how these resources are provisioned (e.g.
fixed or mobile test-beds, physical or virtual resources)
Methods - Cloudification
• Enables large scale integration - scalability
• Provides network functionalities “as a Service” (e.g.
Testbed as a Service)
• Merges IoT with other emerging paradigms of the Future
Internet (e.g. Semantic Web, Cloud Computing, etc)
A Suggested Approach
• A taxonomy of test-beds. For each class, we define
cloudification prerequisites
• Goal: individual test-beds to be organized in a metatestbed platform
• A single application layer accessing and managing
resources from all test-beds (access rights, reputation
and trust mechanisms)
D. Potential themes for EU-Japan Collaboration
Themes:
• Sensor Networks
• IoT
• Distributed Robotics
• Social Networking
Application context:
• Green/smart buildings
• Smart Cities
• Smart e-Health
• Smart Grid