Space Efficiency + Spectral Efficiency = MAZE OF TINY, CONNECTED THINGS Trend: more and more widespread sensing and monitoring data available
an example: the car DATA / INFORMATION OVERLOADHUMAN MACHINE MACHINE HUMAN Trend: much more data than we can cope with
DATA / INFORMATION OVERLOAD, BUT... siloed and bespoke IoT applications APPS APPS APPS APPS APPS APPS APPS APPS APPS PATIENT PATIENT FRIDGE HOUSE APPS PATIENT CAR APPS PATIENT PATIENT PATIENT PATIENT TRUCK SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS
IoT innovation potential...HUMAN MACHINE “Innovation”: one can focus on apps!!! MACHINE HUMAN OBD On Board Diagnostics
IF A WELL-DEFINED INTERFACE INTO CAR SENSORS BRINGS SUCH POTENTIAL... APPS APPS APPS APPS APPS APPS APPS APPS APPS PATIENT PATIENT FRIDGE HOUSE APPS PATIENT CAR APPS PATIENT PATIENT PATIENT PATIENT TRUCK SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS SENSORS
iCore concepts• Virtual Object• Composite Virtual Object• Service / User Level
Virtual Object as OBD across silos To upper iCore levels and Internet VO registry IoT services Semantic VO descriptions Virtual Object SW Agent IoT services Abstraction Proprietary services ICT objects (heterogeneous world) Sensors and actuators Associated physical objects10
what ingredients?• common interfaces to interact with objects (i.e. REST)• + extra containers for metadata• let the systems know what the object is good for, its location (“I am a Temp sensor in Room A”), its accuracy, its energy levels etc.
WHAT ARE VOs GOOD FOR?• OBJECTS REUSE – reuse across different apps, increase availability, hence, increase monitoring / sensing granularity• OBJECTS MGMT – i.e. energy management, contextualised sensing (accuracy vs. sensing frequency) etc.• OBJECTS PROXIMITY – automated selection “by relevance” (see arguments for cognitive technologies in a minute...)Providing IoT systems the ability to self-configure, based on various requirements, and ...
...providing IoT systems the ability to adapt APPS APPS APPS APPS PATIENT FRIDGE HOUSE CAR SENSORS SENSORS SENSORS SENSORS PATIENT is driving the CAR CAR is near the HOUSE PATIENT is near the FRIDGE objects reuse across domains KitchenPresDetect PatientStatusDetect Easy for us...not for a “dumb” computer...
the need for cognitive technologies• iCore Composite Virtual Object (CVO) – aggregation of simple sensing capability – self-maintenance (service maintained in case of failure) increased sensing granularity needed! – System Knowledge • what is available to meet reqs?ability to select alternatives based onwhat metadata we put in the extra VOcontainers “smart but not so much...” CT 1
the need for cognitive technologies• iCore Service Level and overall Cognitive Management FrameworkCT 2,3
DATA / INFORMATION OVERLOAD IF “crash” sight THEN “alertRSA” < distributed sensing sight hearing centralised sensingsmell touch > CT 2
the need for cognitive technologies• factoring “smart logic algorithms” out of developers concerns – IF “crash” THEN “alertRSA” – “crash” (IF VO_x = TRUE THEN crash := TRUE) – (IF VO_x = TRUE AND VO_y = TRUE THEN crash := TRUE)• iCore community: foster “ready meals” for IoT apps factor out cognitive technologies VO_x TAG: IF VO_x = TRUE crash THEN crash := TRUE detect IF (VO_x = TRUE) AND (VO_y = TRUE) VO_y THEN crash := TRUE TAG: IF (VO_x > TH_x) AND (VO_y > TH_y) crash CT 2 detect THEN crash := TRUE
the need for cognitive technologies• rather than for the selection of appropriate templates, here focus is on refinement of selected one according to observed system-reality matching• Real-World-Knowledge “growing” VO_x REFINE TAG: TH_x, and TH_y crash detect assess IF (VO_x > TH_x) AND (VO_y > TH_y) QUALITY of VO_y THEN crash := TRUE PREDICTION TAG: crashCT 3 detect
Summary iCore and Cognitive Technologies• CVO Level “system knowledge” – SLA-driven VO selection / maintainance CT 1 – semantic enrichment semantic-based reasoning – selection by relevance to the needs of the application• deal with data / information overload – template select CT 2 – given VO / CVO “types” find best algorithms that combine these for desired output• deal with data / information overload – learn and predict – given an algorithm, tweak parameters to better align iCore system behaviour to the observed real situation – Real World Knowledge (RWK) “growing” CT 3
main envisaged applications of iCore results• smart-cities and IoT-based monitoring [REF1] P. Vlacheas, R. Giaffreda et al. "Enabling Smart Cities Through a Cognitive Management Framework for the Internet of Things“, to appear in IEEE Communications Magazine - Special Issue on Smart Cities (June 2013) [REF2] 7th May 2013 The Sensing Smart City workshop
The Internet of Things evolution timelineYESTERDAY TODAY TOMORROWThe Dumb IoT The Craft IoT The Cognitive IoT Bear with us, we are building it!
thank you!iCore Website ID Cardwww.iot-icore.eu 3 yrs EU FP7 Integrated ProjectContacts: (started 1st Oct 2011) 20 Partners with strong industrialRaffaele Giaffreda firstname.lastname@example.org 8.7mEur EU Funding EU + China and JapanAbdur Rahimabdur.email@example.com Japan