Deterministic Robotic E-Skin Networks

Today, Ethernet and Time-Sensitive Networking (TSN) are increasingly emerging as key technologies for robotic communication systems. Similar to developments in the automotive sector, robots are adopting zonal architectures in which switched Ethernet networks provide a deterministic communication backbone connecting sensors, actuators, compute modules, and distributed controllers across different parts of the robot [3]. TSN offers important capabilities for such architectures, including time synchronization, bounded latency, traffic prioritization, and reliable delivery of time-critical data. While Ethernet and TSN are well suited for communication between larger robotic subsystems, extending these principles toward highly distributed e-skin platforms is considerably more challenging. At this scale, conventional Ethernet connectivity becomes impractical due to wiring complexity, connector count, power consumption, footprint, and mechanical constraints imposed by flexible electronics. Nevertheless, it would be beneficial to adopt similar communication principles down to this level and transform these electronic structures into a deterministic distributed network that behaves as a robotic nervous system, supporting synchronized sensing, local reflexes, and predictable data exchange. Such an approach could unify communication across the entire robotic system and provide an alternative to today's fragmented set of local interconnect technologies.

This raises a fundamental research question: how can the deterministic communication principles that make TSN attractive at the robot level be preserved within highly distributed e-skin networks? New communication architectures are required that bridge the gap between the robot-wide Ethernet backbone and the dense networks of sensors and embedded intelligence within e-skin platforms. Such architectures must efficiently support distributed sensing, local processing, synchronization, and sub-millisecond reflex actions while remaining compatible with the constraints of flexible electronics and large-area integration technologies.

This PhD will investigate whether deterministic networking principles inside a robot can be extended down to the distributed robotic nervous systems implemented on flexible/stretchable electronic substrates. Particular attention will be given to the physical realization of such networks on flexible/stretchable electronics, including design, material stack selection and fabrication, electromagnetic modelling, mechanical modelling, signal integrity analysis, and communication channel characterization to understand the impact of large-area interconnects, flexible/stretchable substrates, and dense sensor integration on deterministic communication performance. The research will combine flexible/stretchable electronics design, flex-PCB design, electromagnetic and communication-channel modelling, network architecture design, embedded processing, and experimental validation through robotic e-skin demonstrators. The work will closely interact with imec's activities on flexible electronics, large-area electronics, robotic e-skin technologies, and deterministic networking.

• Analyze the communication, synchronization, power distribution, and integration requirements of future robotic e-skins comprising distributed sensing, local processing, and actuation elements embedded in flexible electronic substrates.
• Design a hierarchical robotic nervous-system architecture that seamlessly connects robot-wide Ethernet TSN networks with dense e-skin sensing and actuation fabrics.
• Develop novel communication architectures and protocols that provide deterministic latency, synchronization, and scalability within and across e-skin tiles while respecting the constraints of flexible electronics.
• Investigate the physical-layer realization of distributed e-skin networks through electromagnetic modelling, signal-integrity analysis, communication-channel characterization, and power-distribution network design for large-area flexible electronic substrates.
• Design and optimize flex-PCB and flexible/stretchable electronic architectures integrating sensors, communication interfaces, embedded processing, and power delivery on conformable substrates.
• Investigate communication-compute co-design approaches enabling local reflexes, distributed intelligence, and efficient aggregation of multimodal sensor data.
• Experimentally validate communication architectures and electronic designs through prototype e-skin tiles and multi-tile demonstrators integrated with Ethernet TSN robotic backbones.

• A novel flex-PCB architecture for deterministic robotic nervous systems.
• Communication protocols and synchronization mechanisms supporting deterministic operation, local reflexes, and scalable integration of large numbers of sensing nodes.
• Flex-PCB and flexible/stretchable electronic designs integrating sensing, communication, processing, and power distribution in conformable e-skin platforms.
• Design guidelines and models for scalable e-skin electronics, including routing strategies, tile architectures, connector solutions, and reliability under mechanical deformation.
• Prototype e-skin tiles demonstrating integrated sensing, local intelligence, deterministic communication, and robot-level interoperability.
• Experimental evaluation of performance, latency, synchronization accuracy, energy consumption, scalability, and mechanical robustness of the proposed architectures.
• Scientific publications, demonstrators, and intellectual property at the intersection of flexible electronics, deterministic networking, and robotic sensing systems.