Company or Institution
University of Leeds
Alternative Energy Production & Services
Sustainable Development Goals (SDGs)
SDG 7: Affordable and Clean Energy
SDG 9: Industry, Innovation and Infrastructure
SDG 10: Reduced Inequality
SDG 11: Sustainable Cities and Communities
SDG 12: Responsible Consumption and Production
General description of the AI solution
This project proposes a novel AI solution to digitally assist and scale up coordination of human collective decisions in Smart Cities. Long-standing problems such as power blackouts, traffic jams or over-crowded parking spaces have unprecedented impact on environment and society. They are often a result of coordination deficit: humans failing to align their power consumption, traffic routes or parking space choices at critical moments of need. What if there was a scalable, automated way for citizens to exchange timely information to discover a collective arrangement of their actions that is more sustainable for the environment? What if this collective arrangement could be low-cost, but also fair, inclusive and made with a trustworthy solution? This is the AI solution of collective learning (EPOS project). Via collective learning running on connected personal devices, e.g. smart phones, citizens receive decision-support feedback, e.g. recommendations, calculated as a result of decentralized and privacy-preserving learning interactions between their devices. These optimized decisions may include the moment to turn-off a home appliance to reduce energy, the charging slot of an electric vehicle or the bike sharing station to use. What makes the difference in these optimized decisions altogether is that they can load-balance and improve the efficiency, while meeting a broad spectrum of sustainable development goals. Collective learning comes in stark contrast to mainstream supervised AI solutions that require training with massive personal data and cannot easily scale. Collective learning comes with a well-documented and open-source TRL-6 implementation. It is backed up by several years of cutting-edge research, including an international network of partners, high-profile publication output and open data. Proof-of-concepts in the domain of energy, transport and sharing economies demonstrate its impact on sustainable development. It is currently supported within a £1.4M UKRI Future Leaders Fellowship and a prominent network of partners in industry and cities.
Github, open data repository, prototype or working demo
1] Mashlakov, A., Pournaras, E., Nardelli, P.H. and Honkapuro, S., 2021. Decentralized cooperative scheduling of prosumer flexibility under forecast uncertainties. Applied Energy, 290, p.116706.
 Nezami, Z., Zamanifar, K., Djemame, K. and Pournaras, E., 2021. Decentralized edge-to-cloud load balancing: Service placement for the Internet of Things. IEEE Access, 9, pp.64983-65000.
 Fanitabasi, F., Gaere, E. and Pournaras, E., 2020. A self-integration testbed for decentralized socio-technical systems. Future Generation Computer Systems, 113, pp.541-555.
 Pournaras, E., 2020, August. Collective learning: A 10-year odyssey to human-centered distributed intelligence. In 2020 IEEE International Conference on Autonomic Computing and Self-Organizing Systems (ACSOS) (pp. 205-214). IEEE.
 Pournaras, E., Pilgerstorfer, P. and Asikis, T., 2018. Decentralized collective learning for self-managed sharing economies. ACM Transactions on Autonomous and Adaptive Systems (TAAS), 13(2), pp.1-33.
 Pournaras, E., Yao, M. and Helbing, D., 2017. Self-regulating supply–demand systems. Future Generation Computer Systems, 76, pp.73-91.