Today’s energy requirements for artificial intelligence (AI) are skyrocketing. Considering sustainability constraints, this calls for innovative solutions for energy awareness. Additionally, applications for AI at the edge, such as smart IoT devices, wearable health monitors, and autonomous drones, which perform computations close to where data is generated or collected, are rapidly expanding in the market. These devices require a high level of energy efficiency to enable AI deployment at the edge.

However, energy-aware AI systems will only be unlocked through transdisciplinary co-innovation across the stack, merging tools for monitoring, green software techniques, efficient algorithms, and low-footprint hardware, to achieve the goal for order-of-magnitude energy savings.

This integrative approach for Energy-aware AI Systems is described in more detail in the TU Delft white paper Green AI: Energy-aware Smart Systems from the Edge to the Cloud

12.30 – 13.15 Lunch
13.15 – 13.30 Opening and Introduction
13.30 – 14.30 Inspirational Talks + Discussion

• • Software & algorithms – Przemysław Pawełczak
  • Hardware – Rajendra Bishnoi / Charlotte Frenkel
  • Development & deployment tools and user interface – Aaron Ding

14.30 – 15.00 Coffee Break
15.00 – 16.15 Breakout sessions per topic (for each of the three topics)
16.45 – 17.30 Drinks

• Charlotte Frenkel, Assistant Professor in digital and neuromorphic IC design at TU Delft, Department of Microelectronics of the EEMCS faculty.
  
• Aaron Ding, Senior Associate Professor in Edge AI solutions for cyber-physical systems at TU Delft, Department of Information and Communication Technology of the TPM faculty.
• Przemysław Pawełczak, Associate Professor at TU Delft, within the Embedded Systems Group of the EEMCS faculty, leading sustainable systems lab.
• Rajendra Bishnoi, Assistant Professor in analog neuromorphic computing at TU Delft, within the Computer Engineering (CE) group of the EEMCS faculty.

• Reflection from industry on the whitepaper Green AI: Energy-aware Smart Systems from the Edge to the Cloud
• Challenges and key questions from industry
• How could these challenges be tackled & who should be involved?
• Fostering new collaborations between industry & academics (including funding opportunities)