positions

Task: Battery-free Embedded AI (WP1)

Host institution: TU Darmstadt

Country: Germany

Supervisor: Prof. M. Zimmerling [TU Darmstadt]

Co-suopervisors: Prof. K. Yildirim [UNITN]; Dr. P. Sommer [ABB]

Objectives: (1) To propose formal modelling to dissect and understand the complex interplay of the dynamic energy-harvesting conditions, the capabilities of hardware architecture, and the properties of intermittent software runtime; (ii) To co-design reconfigurable software, hardware, and harvesting architectures; (iii) To validate the proposed solutions on real systems.

Expected Results: 1) Formal modelling of interplay of dynamic energy harvesting conditions, hardware capabilities, and intermittent software; 2) A reconfigurable dataflow architecture combined with a novel algorithm tunes key parameters of a battery-free system to improve
efficiency by more than 1000x over SOTA; 3) Validation on a real proof-of-concept implementation.

PhD enrolment: Doctoral School of TU Darmstadt

Planned secondments: 

• UNITN (3 months, M16-M18): Reconfigurable software & harvesting architecture for battery-free devices, with Prof. K. Yildirim (KPI: joint paper)

• SFI (4 months, M28-M31): Testbed implementation for zero-power smart-farming application (KPI: joint paper)

Candidate profile: computer or electrical engineering, computer science (in order of preference)

Desirable skills/interests: embedded systems (e.g., low-level programming in C/C++, low-power design), signal processing, linear algebra, probability theory, applied optimization, machine learning (e.g., neural network training, TinyML frameworks) (the applicant should be proficient in at least two of the skills)

Application Deadline: February 14, 2025, AoE

Descriptions of all the 18 DC Positions

Submit Your Application HERE