Energy-aware Joint Orchestration of 5G and Robots: Experimental Testbed and Field Validation

Milan Groshev 1 , Lanfranco Zanzi 2 , Carmen Delgado 3 , Xi Li 2 , Antonio de la Oliva 1 , Xavier Costa-Pérez 2 3 4 · 2025

Read Energy-aware Joint Orchestration of 5G and Robots: Experimental Testbed and Field Validation by Milan Groshev 1 , Lanfranco Zanzi 2 , Carmen Delgado 3 , Xi Li 2 , Antonio de la Oliva 1 , Xavier Costa-Pérez 2 3 4 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-19613.txt.

Introduction

Milan Groshev and colleagues field-test OROS, a system that jointly runs 5G networks and teams of mobile robots with an eye on energy. Robots sent outdoors on 5G can offload work to the cloud, yet battery life still caps their missions. OROS coordinates robot navigation and sensing while tuning the cloud services behind them, guided by live feedback.

Trials ran on shop-bought robots over a campus 5G network. By shifting heavy computing to the edge and switching off idle sensors, OROS saved about fifteen percent of robot energy against leading methods. Longer battery life means longer missions and better use of each robot.

Catalog record

CREATORMilan Groshev 1 , Lanfranco Zanzi 2 , Carmen Delgado 3 , Xi Li 2 , Antonio de la Oliva 1 , Xavier Costa-Pérez 2 3 4
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-19613

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Energy-aware Joint Orchestration of 5G and Robots: Experimental Testbed and Field Validation Milan Groshev 1 , Lanfranco Zanzi 2 , Carmen Delgado 3 , Xi Li 2 , Antonio de la Oliva 1 , Xavier Costa-Pérez 2 3 4 Abstract 5G mobile networks introduce a new dimension for connecting and operating mobile robots in outdoor environments, leveraging cloud-native and offloading features of 5G networks to enable fully flexible and collaborative cloud robot operations. However, the limited battery life of robots remains a significant obstacle to their effective adoption in real-world exploration scenarios. This paper explores, via field experiments, the potential energy-saving gains of OROS , a joint orchestration of 5G and Robot Operating System (ROS) that coordinates multiple 5G-connected robots both in terms of navigation and sensing, as well as optimizes their cloud-native service resource utilization while minimizing total resource and energy consumption on the robots based on real-time feedback. We designed, implemented and evaluated our proposed OROS in an experimental testbed composed of commercial off-the-shelf robots and a local 5G infrastructure deployed on a campus. The experimental results demonstrated that OROS significantly outperforms state-of-the-art approaches in terms of energy savings by offloading demanding computational tasks to the 5G edge infrastructure and dynamic energy management of on-board sensors (e.g., switching them off when they are not needed). This strategy achieves approximately ∼ 15 % \sim 15\% energy savings on the robots, thereby extending battery life, which in turn allows for longer operating times and better resource utilization. I Introduction The success of outdoor Search and Rescue (SAR) missions depends on the timely coordination of rescue teams on-site to overcome difficulties in locating missing persons and ensuring a prompt response . To minimize rescue teams’ exposure to hazardous outdoor environments and enhance search and rescue efficiency, there is an increasing demand for deploying coordinated autonomous robots in mission-critical operations [ 1 ] . These robots are equipped with heterogeneous sensors and communication capabilities to explore the environment, determine their location, and search for targets or individuals while providing operators with real-time monitoring information. Additionally, Artificial Intelligence (AI) can aid these autonomous systems by enabling more accurate and adaptive decisions based on real-time multi-sensor data streams [ 2 ] . In this context, the growing demand for autonomous and coordinated robots is driving a technological shift from standalone robot deployments towards connected autonomous robot platforms. The ubiquitous connectivity, high bandwidth and low latency access offered by the 5 th generation of mobile networks (5G), combined with the ability to leverage edge and cloud computing for processing and analytics, enable unprecedented flexibility in deploying modern robotic applications for outdoor scenarios [ 3 ] . Importantly, SAR missions often rely on private 5G networks tailored to the specific requirements of the mission [ 4 ] , [ 5 ] . This approach addresses the unique challenges of SAR sites, which are frequently located in remote or rural areas—such as mountainous regions—where public 5G coverage is typically unavailable. Private 5G networks ensure that all network resources are exclusively dedicated to the mission, minimizing the risk of resource contention or external fluctuations while providing reliable communication for mission-critical robotic operations. However, a trade-off arises when pursuing higher degrees of robot autonomy and coordination. While mobility remains the primary source of energy consumption [ 6 ] , the computing and processing of multi-sensor data required for outdoor exploration use cases further increase the energy demands of robotic platforms [ 7 ] . Robot operations are heavily constrained by their hardware’s computational capabilities and associated energetic aspects. Currently , most of the real-world learning problems faced by robots can only be addressed by offloading dense data to more powerful computing infrastructure or dedicated edge/cloud computing units [ 8 ] [ 9 ] . Therefore, ensuring robust, high-speed, and low-latency communication on demand is crucial . The current state-of-the-art is evolving by focusing on innovative solutions in individual domains. For example, advancements in the network domain have led to high-speed, low-latency communication, but result in suboptimal robotic performances due to insufficient integration and real-time awareness. Conversely, developments in the robotic domain enhance autonomy and functionality, but often neglect the network aspect of the end-to-end system, limiting real-time communication and coordination capabilities in advanced use cases. Motivated by the need for joint orchestration of the robotics and network domain, our previous work [ 10 ] introduced OROS , an orchestration framework that uses the knowledge and requirements obtained from the robots ( robot-as-a-sensor ) to seamlessly optimize the network resources and robotic operations. While OROS was extensively evaluated through simulations [ 11 ] , it has yet to be validated in real-world scenarios. Such field validation is essential to fully characterize its potential, implementation challenges and limitations. To address this gap , this paper focuses on the experimental evaluation of OROS in an outdoor testbed with real mobile robots and local 5G connectivity. Our implementation provides a comprehensive testing environment for experimenting with various network, computation, and robotic configurations. Using this testbed, we deploy open-source mobile robots, an object detection robotic service, and a prototype of OROS to optimize robot trajectories and minimize energy consumption. The main contributions that distinguish this work from the state - of - the - art are: • We perform a comprehensive energy consumption profiling of mobile robot applications. • We develop a cloud-based, containerized cross-field testbed that serves as a playground for testing and validating innovative solutions integrating both network and robotics domains. This testbed also incorporates 5G technology and features dedicated APIs specifically developed for seamless interaction between the network and robotic components. • We build a digital model relying on the ROS-based Gazebo s imulator to emulate the behavior of multiple mobile robots into a digital representation of the deployment environment to facilitate testing and fine-tuning of OROS before its deployment in the physical environments. • We design and implement OROS in a real outdoor testbed, integrating both 5G and robot domains, evaluating its performances through a field trial and demonstrating its feasibility in realistic environments. The software and APIs of OROS , the digital model of the test environment, as well as the measurement data from the testbed will be available to the public after acceptance of the paper. This can provide to both the networking and robotic research community a playground for testing and evaluating the proposed solution with real field experiments, in turn fostering openness in this research field and encouraging collaboration and knowledge-sharing. The remainder of this paper is structured as follows. Sec. II presents the needed background on 5G and robot orchestration as well as a short description of OROS and its interaction with the different modules. Sec. III introduces the real-world Proof-of-Concept (P o C) testbed and details its main components. Sec. IV provides an exhaustive energy profiling based on measurements from two 5G-enabled mobile robots. Sec. V validates the design principles of our solution and provides a field test validation, highlighting the main benefits derived from our approach. Sec. VI discusses the algorithm complexity, providing insights on improving the overall system scalability. Sec. VII summarizes related works in the field. Finally, Sec. VIII concludes this paper. Fig. 1 : Overview of the architectural building blocks.

Read the complete text