Marco Levorato

Papers

Total Citations

H-Index

About

Marco Levorato is a leading researcher at the forefront of edge intelligence and distributed computing for the Internet of Things (IoT). His work fundamentally addresses the tension between deploying powerful Deep Neural Networks (DNNs) on resource-constrained mobile devices and the demands of safety-critical applications. Levorato’s primary contributions lie in the paradigm of Split Computing (SC), where he has pioneered methods to partition DNNs between mobile devices and network servers. His highly cited 2024 paper, "Evaluating the Reliability of Supervised Compression for Split Computing," with 8 citations, provides a critical framework for assessing the trade-offs between compression and inference accuracy in this emerging field. He further advanced this line of inquiry in "Enhancing the Reliability of Split Computing Deep Neural Networks," directly tackling the robustness challenges for autonomous systems and biomedical robotics. By systematically addressing the computational bottlenecks of modern AI, Levorato is shaping the future of how intelligent applications can operate reliably at the edge, enabling a new generation of responsive and trustworthy IoT systems.