Octavian M. Machidon

Octavian M. Machidon

Assistant Professor, University of Ljubljana
Human-Centered AI • Edge & Adaptive AI Systems • Ethics and Anthropology of Artificial Intelligence

Current projects (Principal Investigator)

CARDIO-FL — Context-Aware Cardiac Monitoring with On-Device AI and Federated Learning

Role: Principal Investigator and Lead Partner (University of Ljubljana, Faculty of Computer and Information Science)
Funding: European Commission — ENFIELD Innovation Scheme
Duration: January 2026 – June 2026

Aim. CARDIO-FL develops a chest-mounted wearable system for real-time cardiac anomaly detection by fusing ECG, PPG, and smartphone inertial data. The project focuses on context-aware, privacy-preserving monitoring through on-device, self-supervised, and federated learning, targeting a validated prototype at TRL 4–5.

  • Self-supervised and federated learning for personalized cardiac monitoring
  • Multimodal signal fusion (ECG, PPG, IMU) and adaptive inference pipelines
  • Smartphone integration, optimization, and system-level evaluation
  • Validation against commercial and clinical cardiac monitoring devices

J6-60105 — Theology and Digitalization: Anthropological and Ethical Challenges

Role: Principal Investigator for UL FRI (institutional PI) (University of Ljubljana, Faculty of Computer and Information Science)
Funding: Slovenian Research Agency (ARRS)
Duration: 2025–2027

Aim. This interdisciplinary project examines how digitalization and artificial intelligence reshape human identity, embodiment, empathy, spirituality, and human dignity. Drawing on theology, philosophy, psychology, and computer science, the project develops a theological and ethical anthropology suitable for the digital age.

  • Digital identity, embodiment, and vulnerability in AI-mediated environments
  • Digital communication, empathy, and human relationships
  • Religious experience and spirituality in digital culture
  • Ethical guidelines for responsible, human-centered digital technologies

Selected outputs:

  • Machidon, O. M. (2025). Beyond Algorethics: Addressing the Ethical and Anthropological Challenges of AI Recommender Systems. Journal of Media Ethics. DOI
  • Machidon, O. M. (2025). Forming Hearts and Minds: Challenges and Renewal in Catholic Education in the Age of AI. International Studies in Catholic Education. DOI
  • Machidon, O. M. (2025). Theology in Dialogue with Artificial Intelligence. In K. Glombik (Ed.), Theology in Dialogue: An Interdisciplinary Perspective on Contemporary Issues (pp. 127–145). Opole. DOI
  • Machidon, O. M. (2024). Upholding Human Dignity in AI: Advocating Moral Reasoning over Consensus Ethics for Value Alignment. Zagadnienia Filozoficzne w Nauce, 77, 25–39. Journal

Completed projects (Principal Investigator)

AgriAdapt — Energy-Efficient Agriculture Using UAVs and Real-Time Neural Network Adaptation

Role: Principal Investigator (University of Ljubljana, Faculty of Computer and Information Science)
Funding: European Commission
Duration: January 2023 – October 2023

Aim. AgriAdapt developed a dynamic, resource-efficient machine learning framework for precision agriculture using UAV imagery, enabling real-time neural network adaptation and significant reductions in on-board energy consumption while preserving weed-detection accuracy.

  • Context-aware neural network adaptation and dynamic model slimming
  • Real-time UAV-based weed detection from aerial imagery
  • Energy-efficient deployment on low-cost UAV platforms
  • Technology validation and groundwork for commercialization

Selected outputs:

  • Machidon, O. M. et al. (2023). AgriAdapt: Towards Resource-Efficient UAV Weed Detection Using Adaptable Deep Learning. NET4us ’23. ACM DL
  • Igret, I. C., Machidon, A. L., & Machidon, O. M. (2024). Context-Based Adaptation of Neural Network Compression for UAV Weed Detection. IEEE ICCP 2024. IEEE Xplore
  • Machidon, A. L., Krašovec, A., Pejović, V., & Machidon, O. M. (2025). SqueezeSlimU-Net: An Adaptive and Efficient Segmentation Architecture for Real-Time UAV Weed Detection. IEEE JSTARS. IEEE Xplore
  • Machidon, A. L. et al. (2025). A Low-Cost UAV System and Dataset for Real-Time Weed Detection in Salad Crops. Electronics, 14(20), 4082. Open access

Embodied Conversational Agent with Avatar Interface (Natural Language)

Role: Principal Investigator
Funding: National Cultural Fund Administration (Romania)
Duration: 2019

Aim. Development of an embodied, avatar-based conversational agent with multimodal interaction and dialogue management, deployed as a virtual guide in a museum environment.

  • Multimodal conversational AI design
  • Dialogue management and avatar integration
  • Deployment and evaluation in a cultural heritage setting

Selected outputs:

  • Duguleană, M. et al. (2020). A Virtual Assistant for Natural Interactions in Museums. Sustainability, 12(17), 6958. Open access

MULTICOLOR — Multispectral Imaging for Multi-Color Image Acquisition

Role: Principal Investigator
Funding: Transilvania University of Brașov
Duration: 2017–2018

Aim. Development of a multispectral imaging system for multi-channel image acquisition and processing, targeting applications in cultural heritage analysis, material characterization, and environmental monitoring.

  • Multispectral acquisition hardware and pipelines
  • Calibration and signal processing for multi-channel imaging
  • Applications in heritage, materials, and environmental monitoring

Selected outputs:

  • Man, C., Machidon, O. M., & Ivanovici, M. (2018). VIPERA 1.0 – A Versatile Imaging Platform for Education, Research and Applications. ISETC 2018. IEEE Xplore