Robin Ohs

• 2026Dark Clouds Rising in Low-Earth Orbit: On Environmental Limits to Massive Orbital AI · LEO-NET 2026
• 2025Dirty Bits in Low-Earth Orbit: The Carbon Footprint of Launching Computers · HotCarbon'25
• 2025PhantomLink: Emulating Virtual End-to-End Links on Ground and in Orbit · ANRW'25

• Sep 2026
  Data centers : pourquoi l’IA pousse l’industrie à conquérir l’espace ?

  Sciences et Avenirabout Dirty Bits

• Dec 2025
  Space-Based Data Centers Could Power AI with Solar Energy—At a Cost

  Scientific Americanabout Dirty Bits

1. 2026

   Dark Clouds Rising in Low-Earth Orbit: On Environmental Limits to Massive Orbital AI

   Robin Ohs*, Gregory F. Stock*, Andreas Schmidt, Juan A. Fraire, Jörg Ott, Holger Hermanns

   LEO-NET 2026WorkshopDOICiteSlides

   * authors contributed equally

   ›Abstract

   In 2026, we saw rising numbers of proposals for orbital data centers to facilitate AI, e.g., by SpaceX (up to one million satellites), Blue Origin, and Google–yet none include rigorous lifecycle sustainability analysis. We present ESpaS-ODC, a lifecycle carbon model that accounts for the three subsystems physically unavoidable at GPU-class power densities but absent from prior work: a thermal radiator sized from ISS data, solar-array degradation and eclipse margin, and cold-standby spares for no-repair access. Applying the model to an edge data center in a 510 km orbit reveals that the service-overhead-scaled radiator weighs about eleven times the GPU it serves and that modeled power and thermal infrastructure dominate launched component mass–not compute. Using mission duration T as the denominator to amortize launch, re-entry, and manufacturing carbon, model-boundary parity with a global-average terrestrial DC occurs within the first two mission years in our parameter sweep. With respect to a renewables-powered green-energy DC (Finland grid intensity: 68 gCO2e/kWh, PUE: 1.2), parity requires multi-year missions with cold-standby spares compensating the shorter component lifetimes. Idealized dawn-dusk SSO (no-eclipse regime, β > βcrit) eliminates the eclipse battery, cutting modeled component mass significantly (e.g., 25% on the 1 kW-ODC for a 3 yr mission), but leaves the radiator unchanged: high-beta orbits solve the battery problem, not the thermal problem. Finally, carrying one full cold spare (r = 2) increases amortized carbon per GPU hour by 40% on Starship and 34% on Falcon-9 at T = 3 yr, while its dependability benefit remains to be quantified.

2. 2025

   Dirty Bits in Low-Earth Orbit: The Carbon Footprint of Launching Computers

   Robin Ohs, Gregory F. Stock, Andreas Schmidt, Juan A. Fraire, Holger Hermanns

   HotCarbon'25WorkshopDOICite

   ›Abstract

   Low-Earth Orbit (LEO) satellites are increasingly proposed for communication and in-orbit computing, achieving low-latency global services. However, their sustainability remains largely unexamined. This paper investigates the carbon footprint of computing in space, focusing on lifecycle emissions from launch over orbital operation to re-entry. We present ESpaS, a lightweight tool for estimating carbon intensities across CPU usage, memory, and networking in orbital vs. terrestrial settings. Three worked examples compare (i) launch technologies (state-of-the-art rocket vs. potential next generation) and (ii) operational emissions of data center workloads in orbit and on the ground. Results show that, even under optimistic assumptions, in-orbit systems incur significantly higher carbon costs–up to an order of magnitude more than terrestrial equivalents–primarily due to embodied emissions from launch and re-entry. Our findings advocate for carbon-aware design principles and regulatory oversight in developing sustainable digital infrastructure in orbit.

3. 2025

   PhantomLink: Emulating Virtual End-to-End Links on Ground and in Orbit

   Robin Ohs*, Gregory F. Stock*, Juan A. Fraire, Holger Hermanns, Andreas Schmidt

   ANRW'25WorkshopDOICite

   * authors contributed equally

   ›Abstract

   Low-Earth Orbit (LEO) networks are highly dynamic due to their closeness to Earth and fast movements. At the same time, these dynamics are also highly predictable as the LEO satellites have their fixed orbits. As this predictability is rarely found in terrestrial networks, existing network evaluation tools cannot be used in associated Internet research.We propose PhantomLink, a tool to emulate an evolving virtual end-to-end (E2E) link in real time using the unmodified Linux networking stack. PhantomLink emulates general scenarios—which are inspired by, but not limited to, LEO networks. The tool represents a virtual network interface that can be used to send/receive packets as if they were using an actual multi-hop path. Finally, we showcase our tool in the context of current orbital Internet research.

4. 2025

   FLoRaSat 2: Simulating Cross-Linked Direct-to-Satellite IoT LEO Constellations

   Alexander Y. Choquenaira-Florez, Robin Ohs, Juan A. Fraire, Hervé Rivano

   ASMS'25ConferenceDOICite

   ›Abstract

   Direct-to-Satellite IoT (DTS-IOT) represents a promising solution for data transmission in remote regions where terrestrial infrastructure deployment is unfeasible. In DTS-IOT scenarios, Low-Earth Orbit (LEO) satellites function as in-orbit gateways. Addressing the need for practical simulation tools, we present FLORASAT 2, an open-source, event-driven, end-to-end simulation tool leveraging OMNET++. The original FLO-RASAT met many DTS-IOT Medium Access Control (MAC) requirements. Still, this enhanced version introduces advanced Inter-Satellite Link (ISL) communication modules, including a helper for constellation creation, dynamic ISL topology control, routing, and analytics, facilitating the thorough evaluation of constellation-grade DTS-IOT networks. These new features allow detailed simulation scenario configuration, flexible support for developing and including diverse routing algorithms, and the tooling to perform automated data analysis from parametric simulations. Overall, the simulator enables the analysis of complex behaviors in DTS-IOT environments, optimizing performance and enhancing connectivity and efficiency in large-scale satellite IoT constellation networks.

5. 2025

   HACKATHON: Carbond

   Robin Ohs, all authors of the Dagstuhl Seminar 24351

   Dagstuhl Seminar 24351SeminarDOICiteSlides

6. 2024

   Dirty Electrons: On the Carbon Intensity of Stored Energy

   Robin Ohs, Henry Janson, Andreas Schmidt, Luis Gerhorst, Benedict Herzog, Timo Hönig

   IGSC'24ConferenceDOICiteSlides

   ›Abstract

   To reduce carbon emissions, system software must accurately trace and properly attribute these emissions to applications that cause energy consumption and hardware degradation. While the carbon intensity of the grid energy is readily available from online services, it is often overlooked that computing systems do not directly and immediately consume this energy. Instead, energy coming from the grid is first transformed by power supplies (PSUs) and then often stored in batteries. In this work, we investigate how PSUs and batteries affect the carbon intensity of the energy that actually powers the computer system. We analyze the extent to which specific hardware characteristics need to be taken into account in order to accurately attribute operational and embodied emissions to applications. To do this, we propose an analytical model for the carbon intensity of stored energy, which includes power supply and charging efficiency (which inflates operational emissions) and battery degradation (which causes embodied emissions). We conclude with a look at the future of system-level management of emissions caused by stored energy.

7. 2024

   Quantitative analysis of segmented satellite network architectures: A maritime surveillance case study

   Juan A. Fraire, Santiago Henn, Gregory Stock, Robin Ohs, Holger Hermanns, Felix Walter, Lynn Van Broock, Gabriel Ruffini, Federico Machado, Pablo Serratti, Jose Relloso

   ComNetJournalDOICite

   ›Abstract

   This paper presents an in-depth trade-off analysis of a Swarm Satellite Constellation (SSC) Mission for Earth observation that leverages Segmented Architecture (SA), a concept designed by the Argentinian Space Agency (CONAE) within the New Space philosophy. This architecture consists of a scenario featuring a networked constellation of small, cooperative satellites to enhance mission flexibility, reliability, coverage, and cost-effectiveness. Despite its promising prospects, SA features challenges in its mission design and definition phases due to the complex interplay between distributed space systems, technological innovation, and geographical landscapes. Our study analyzes an innovative quantitative analysis framework integrated with Ansys' Systems Toolkit (STK). The resulting software tool models critical components, including ground and space segments, orbital dynamics, coverage, onboard processing, and communication links. We focus on a hypothetical SARE mission to detect illicit maritime activity near Argentina's Exclusive Economic Zone (EEZ). This case study constitutes an archetypal mission elucidating the architecture's benefits and complexities, addressing swarm coverage, contact dynamics, and data handling strategies. Results contribute to discussions on the practical trade-off in current and future Segmented Satellite Architectures with multiple mission objectives.

8. 2023

   carbond: An Operating-System Daemon for Carbon Awareness

   Andreas Schmidt, Gregory Stock, Robin Ohs, Luis Gerhorst, Benedict Herzog, Timo Hönig

   HotCarbon'23WorkshopDOICite

   ›Abstract

   To reduce the carbon footprint of software, it is imperative that systems first become aware of their footprint. Despite various proposals to make software carbon aware via application-level development kits, we believe awareness of and adjustment to carbon-emission information is an operating-system duty—similar to how it manages time information today. In this paper, we motivate and envision carbond, a Linux-based service to mediate carbon information between hardware (including power supply) and application-level software. Following the recently established Software Carbon Intensity (SCI) standard, carbond deals with operational as well as embodied emissions and has different units of work (low level as CPU cycle up to high-level as user request) in mind. In addition to the service itself, we showcase how it can be used by application SDKs (libraries) as well as command-line utilities.

• ESpaS

  ActiveMaintainer2026

  Estimates the carbon cost of orbital computing and compares on-orbit versus on-ground data processing across mission designs.

  • Orbital Datacenters
  • Space
  • Sustainability

• phantomlink

  ActiveMaintainer2025

  Emulates time-varying end-to-end links with Linux network namespaces, so any application can be tested under changing delay, data rate and routes.

  • Linux
  • Emulation
  • Networking
  • Space

• Carbond

  MaintenanceMaintainer2024

  An operating-system daemon that reports the carbon intensity of a system's energy, accounting for PSU losses and stored battery energy.

  • Linux
  • Daemon
  • Sustainability
  • Carbon

• FLoRaSat

  ActiveAdvisory (Former Dev)2023

  An OMNeT++ framework for simulating LoRa-based satellite IoT networks and routing in LEO constellations.

  • Omnet++
  • Routing
  • Networking
  • Space

1. 2023 – nowPhD student · Saarland University
2. 2023Research Assistant · Saarland University
3. 2021 – 2023MSc Computer Science · Saarland University
4. 2021 – 2022Student Research Assistant · Saarland University
5. 2020 – 2021Developer · German Air Force
6. 2017 – 2020BSc Cybersecurity · Saarland University
7. 2017A-Levels (Abitur) · Gymnasium Saarburg

h-index: 4i10-index: 2Erdős number: 4Dijkstra number: 4