Loading…
Quantify the Unquantifiable: Innovative Methods for Next-Gen Sensing
Abstract: Standard sensing technologies cannot be always suited for to quantify phenomena that cannot be directly measured due to the unavailability of dedicated sensors, or due to a general unfeasibility (e.g., practical constraints, expensiveness of infrastructures, etc.). Enhancing sensing with modern advanced methods can make non-quantitative techniques able to remedy this issue. The Special Session aims to collect papers from the community involved in this topic.
Organizers
Energy-Efficient Edge AI in Resource-Constrained Critical Systems
Abstract: As sensor networks increase in density, transmitting raw data to the cloud becomes unsustainable, which requires a paradigm shift from “passive data collection” to “smart sensing”. Moreover, the deployment of learning models in critical environments requires strict constraints on power, bandwidth, and latency.
This Special Session bridges the gap between sensor hardware constraints and modern machine learning requirements, addressing the intersection of advanced data acquisition strategies and Edge AI to achieve energy autonomy in the Healthcare and Aerospace domains.
We invite contributions focused on innovations that optimize the entire chain, from physical sensing to final decision-making. Key topics include implementing smart data acquisition strategies to reduce sensor duty cycles, applying Edge AI compression techniques suitable for microcontrollers, and identifying trade-offs between model accuracy, energy consumption, and latency. These methodologies are contextualized, but not limited to, within two primary domains: Healthcare, where wearable devices require low-power operation to ensure continuous monitoring and data privacy; and Aerospace, where limited telemetry bandwidth necessitates autonomous onboard processing to guarantee mission reliability.
Organizers
Hyperconnected Human Bodies –Distributed and Smart Onlife Sensing
Abstract: The convergence of sensor and embedded computing technologies is enabling sensor units to process digital workloads directly at the point of data collection. This development facilitates the integration of heterogeneous sensor nodes distributed across the human body, offering a more comprehensive understanding of physiology and movement in real-world conditions. The resulting data creates new opportunities for the detection of biomarkers, continuous monitoring, personalized medicine, early intervention, and human-computer interaction.
Building on this shift toward localized and distributed intelligence, emerging RF bands and protocols such as
next generation WiFi (ie. WiFiHaLow, WiFi 8) and new BLE generations deliver higher throughput with lower energy requirements, while human body communication offers a private and ultra-low power alternative in proximity scenarios. Together, these technological developments create an unprecedented opportunity to rethink how we sense, compute, and communicate around the body in a cloud connected world.
The motivation for this special session is grounded in the scientific relevance of these innovations, as they enable new applications in health, fitness, human performance, and interactive systems.
Organizers
Advanced Sensing for Intelligent Transportation Systems
Abstract: The proposal emphasizes the strategic importance of advanced sensing in intelligent transportation systems, highlighting that the deployment of reliable, distributed, and heterogeneous sensor networks (including cameras, LiDAR, radar, and connected vehicles) enables the accurate estimation of critical variables to support real-time adaptive control algorithms, anticipate congestion, adjust traffic operations within milliseconds, and mitigate incidents before they propagate. The integration of these technologies fosters information fusion and the development of digital twins, which are crucial for ensuring safety, resilience, and energy efficiency under highly variable demand, extreme weather conditions, and the increasing presence of automated vehicles.
The Special Session aims to address key challenges in sensing architectures explicitly designed for real-time adaptive control, including temporal synchronization, interoperability, cybersecurity, calibration, and large-scale maintenance, by promoting a holistic approach to the joint design of sensing, communication, and control strategies. Structured for 90 minutes, the session will feature six oral presentations focused on sensing architecture and its implications for real-time adaptive control, followed by a joint discussion exploring common challenges, future research directions, and opportunities for collaboration. It will conclude with a summary of the session’s main technical contributions.
Organizers
Imaging Sensors in Intelligent Systems
Abstract: Cameras have become one of the predominant sensors inintelligent systems due to their rich data output, non-invasive nature and rapid algorithmic advances. This session addresses the critical challenge of advancing computer vision algorithms—such as detection, classification, tracking, semantic segmentation and 3D reconstruction for direct application in robust, real-world systems.
We seek to showcase and discuss works that simultaneously advances the state-of-the-art in computer vision tasks and demonstrates their important role within applications and integrated intelligent systems.
Our goal is to bring together researchers pushing the boundaries of fundamental vision models with those engineering novel sensor-based solutions. The session will highlight work where the camera is not just a component, but the central source of perceptual intelligence.
The session will allow the symposium's audience to explore how vision centric approaches are providing smart sensor solutions for current and future applications. We hope the attendees will gain insights into vision algorithms that are more efficient, accurate, and capable of operating under real-world constraints.
Organizers
Loading…
Loading…