a Pilot for Energy District Management
Transcription
a Pilot for Energy District Management
INTrEPID Project: a Pilot for Energy District Management C. Borean, R. Drogo De Iacovo, G. Di Bella, Telecom Italia S.p.A. R. Angelucci, RSE S.p.A. 28/09/15 Outline • Project Overview and Pilot focus – goals, innovaJons, design choices, typeseMngs, technologies • Users engagement – touchpoints, interacJons, raising awareness • energy district and consumpJons habits – making data valuable 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 2 INTelligent systems for Energy Prosumer buildIngs at District level FRIGO WIFI SMART PLUG SMART GATEWAY SMART INFO • EU FP7 project on smart energy districts with a real world pilot • BigData IoT plaXorm with inside a machine learning brain, which learns users habits and do its best to match district consumpJon requirements • Mobile app as the main user touchpoint: 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 3 Pilot focus • 55 pilot sites • 36 in Italy, 19 in Denmark (choose in such a way to widen the projects relevance results) • 2 test sites in each conuntry • 35% equipedd with PV systems (equally distributed in Italy and Denmark, but different power peak generaJon) • 2 locaJon equipped with energy storage 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 4 Project Overview • Interoperable Home Area Network technologies: – ZigBee HA 1.2, WiFi, Z-‐Wave, ModBus on IEEE802.15.4 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 5 Human-‐in-‐the-‐loop approach Goals: raise users consumpJon awareness, sJmulate their green-‐behavior, and acJvely parJcipate the district energy opJmizaJon task. EducaRve – Enabling -‐ Empowering • EducaRve: thanks to consumpJon charts related to single appliance, both in mobile app and in themaJc newslefers for inter-‐ users comparisons • Enabling: give users the possibility to measure not plug-‐equiped appliances by means of «energy whatch feature» or turn ON & OFF their appliances • Empowering: influece the system by means of interacJon tools which reinforce algorithm learnings. 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 6 ThemaRc newsleUers Bi-‐weekly INTrEPID newsleUer: the piloters have been periodically updated with bi-‐weekly newslefers reporJng as anonymized informaJon the comparison of exisJng usage in: • each home (Building Unit) and per capita • specific appliances This comparison with other users provides an improved engagement of the users and a key addiJonal level of understanding about their energy behaviors INTrEPID newsleUer structure: • Topic introducJon • Insights and related charts • Tips & News 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 7 ThemaRc newsleUers The core secJon has been focused on consumpJon awareness by means of • charts on specific appliances or house consumpJon as well as energy cycles… • ... but also on emerging usage paferns and “energy personaliJes” 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 8 District core – Learning from habits Measure INTrEPID home System 28/09/15 Habits detecRon INTrEPID Cloud Brain Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 9 District core -‐ ...to change user behavior with proper schedules AcRvate or schedule delay start INTrEPID Cloud Brain Brain iniJally schedules starJng from acJvaJon Jme close to habits GAIN TRUST from the users and maximize acceptance rate Aper this try to MOVE habits to fit desired load profile (tunable parameter wieghJng more users habits or district desired curve) 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 10 SuggesRon response dashboard Response dashboard let the system adapJvely reinforce its learning on users’ habits. 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 11 GamificaRon approach The reward system seems to be a crucial aspect for moving users habits: • +15% in acceptance rate (with regard to a preliminary phase where no credits were given for accepted suggesJon) for the same habits weight in reccomendaJon algorithm 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 12 Thanks, quesJons? 28/09/15 Giuseppe Di Bella, SWARM Joint Open Lab Telecom Italia S.p.A. 13