Stefano Giordano is Associate Professor at the department of Information Engineering (IE) of the University of Pisa. He is also co-founder of the Juniper Networks Higher Learning Center and an active participant in various European projects in the field of ICT. Prof. Giordano will be General co-Chair at Cyclone 2015, the EAI International Conference on CYber physiCaL systems, iOt and sensors Networks (October 26, Rome), co-located with the IOT360 Summit. In our interview, he talks to us about Cyber Physical Systems and his expectations from the forthcoming conference.
As General Chair of Cyclone 2015, would you tell us what expectations are rising on the conference?
The main expectations come from the idea that we want to stimulate the participation in the conference among experts coming from different areas of expertise in the fields of WSAN (Wireless Sensor and Actuator Network), Cyber Physical Systems and IoT. It will be an opportunity to verify how new hardware and software architectures (considering the enabling technologies) meet the (sometimes hard) requirements of certain application areas. It is well known that Wireless Sensor and Actuator Networks composed of hardware and software functionalities will have to be ‘de-verticalized’ thanks to flexible, dynamic, energy efficient, adaptive architectures that could be effective in different application domains ranging from e-health to smart agriculture. To become part of an IoT platform, WSAN will have to be integrated with ‘backhauling networks’ (such as fixed or cellular networks) and to the hardware/software components inside the cloud. Physical sensors, actuators and end-user devices such as Tablets or Smartphones will also play a relevant role in this arena. We hope that at the conference, thanks to experts in each application area, the real needs from each domain could meet adequate technological solutions and not vice-versa where engineers try to solve problems that they don’t know in the inner details. New secure, reliable, manageable, energy efficient, reconfigurable and reprogrammable architecture will have to become the answer to real problems and we hope to contribute with this conference to new progress in the area enabling new steps ahead in the field of open innovation and in the rising of the Internet of the Future.
Cyber Physical Systems and the Internet of Things are often thought as two sides of the same coin. Could you give us an overview of the differences and relations between the two concepts and related systems?
Cyber Physical Systems are a part of the Internet of Things such as Wireless Sensors and Actuator Networks. They can be considered such as end systems, switches/routers for the Internet. The Internet of Things is again a ‘logical view’, an ‘abstraction’, of heterogeneous physical components that, event more than in the past, interact with the cyber representation of those and others systems. The control of single or composite physical systems (from industrial plants, to fluid dynamics experiments, to the human body) represent the most advanced frontier of mutual interaction between a physical space and what it is called a cyberspace. Here specifically the modeling of some of the physical systems in the cyberspace require complex representation such as the ones obtained using differential equations, stochastic modeling, chaotic systems, etc. Any human representation of a complex physical system can become (in one or more computers) a cyber representation of the system itself and this will be the starting point for its control, for the forecasting of some critical behaviors or… just to obtain new forms of arts and creativity.
What are the main challenges connected to a concrete application of the results in this field of research?
Some of the phenomena that we want to measure are extremely critical, the quota of information that they produce is difficult to be extracted, and the actuators and control part can be even more challenging. It is definitely a new Internet with even more critical issues related to privacy, security and reliability aspects. In-network processing, virtualization, software defined approaches (in the sensing, radio and network domains) open new opportunities and challenges. Again, such as for the Internet that we already know, abstract mathematical modeling, simulation will be not enough for the design of the network, and the effectiveness of any proposed solution will have to be validated with an intense experimental activity even on large-scale experiments.
