DC8 — Grid-Aware AI Workload Mobility for the Computing Continuum

Host: Technische Universität Berlin (TU Berlin), Berlin, Germany — Chair of Software and Business Engineering.
Work package: WP2 (Carbon-Aware Intelligence and Control)
Duration and start: 36 months, full-time; start between February and September 2027 (latest start September 2027).

The project

Today’s tools can move services between different computing sites, but they do not understand the power grid, and they struggle to move running AI jobs safely. In this project, you build the missing grid-aware layer for the orchestration of computing resources and AI workloads along the edge-cloud continuum.

You will create middleware that sits on top of existing cloud and edge frameworks and can move stateful AI workloads across the edge-to-cloud continuum, or switch between AI model variants in place, in response to grid signals, while keeping security policies in force during every move.

What you will work on:

• Build middleware for moving AI workloads across the edge-cloud continuum.
• Develop mechanisms for stateful migration of AI workloads and for in-place switching between AI model configurations at runtime.
• Expose one middleware API for workload-mobility commands, with security policy enforced during migration.

What you will deliver:

• Middleware that moves workloads smoothly across different testbeds.
• A transition engine that completes changes within defined time budgets.
• A performance characterization report.

Secondments

You will spend two research visits with partners in the network:

• Deutsche Telekom – T-Labs (Germany) — 3 months from around October 2028 (industry), with Ingo Friese: test workload-migration performance on Deutsche Telekom’s edge infrastructure.
• Lund University (Lund, Sweden) — 4 months from around January 2029 (academic), with Johan Eker: combine the workload-mobility middleware with the intent-driven facilities control.

The host

The Chair of Software and Business Engineering has been established at TU Berlin in 2025, and conducts research at the intersection of distributed systems and business engineering. A particular research focus is on distributed machine learning, edge computing, and distributed trust.

Supervision

Main supervisor: Stefan Schulte — Head of the Chair of Software and Business Engineering, TU Berlin. He works on distributed systems and edge AI.

• ORCID
• Google Scholar
• LinkedIn
• DBLP

Co-supervisors:

• Ingo Friese — Deutsche Telekom (T-Labs)
• Johan Eker (ORCID) — Lund University

Profile we are looking for

• Very good Diploma or Master university degree in Computer Science, Informatics, Data Science, Business Informatics, or a related discipline is expected
• Good scientific communication and writing skills.
• The ability to develop methods, concepts, and models, as well as their realization and evaluation and the willingness to contribute to scientific projects.
• In-depth interest in scientific problems and the motivation for independent and goal-oriented research.
• Knowledge in distributed systems
• Curiosity, initiative, and very good spoken and written English

You must also meet the MSCA eligibility rules — no PhD yet, the mobility rule, and open to all nationalities.

How to apply

Official vacancy at the host institution: link coming soon.

Until then, apply through the central AEGIS application portal: AEGIS application portal (reference position DC8), and read what to include in your application.

Questions about this position? Contact the coordinator, Prof. Paul Pop, aegis@compute.dtu.dk.