Research

This strand concerns the practical deployment of learning-based methods in civil engineering: structural health monitoring, predictive maintenance, optimisation and decision support. The emphasis falls less on novel architectures than on how established techniques, from convolutional and graph neural networks to deep reinforcement learning, can be integrated into engineering workflows whose failure modes carry public consequences.

Intelligent infrastructure is only as trustworthy as the data on which it runs. Work in this area develops governance frameworks, stewardship roles, quality metrics and compliance mechanisms, and examines privacy-preserving techniques such as federated learning and differential privacy as instruments of governance as well as of computation.

A substantial part of the recent monograph programme treats the city as an adaptive system that can anticipate and absorb shocks. The books examine digital twins, predictive disaster modelling, climate adaptation, health infrastructure and data-driven urban design, and draw on documented initiatives in cities across Europe, North America, the Middle East and Asia.

Drawing on more than twenty-five years of running university computing laboratories, at Panteion University in 1998–99 and at NTUA since 2000, this strand addresses digital transformation as an organisational problem as much as a technical one. It includes strategic analysis of Greece's digital transition, e-government interoperability and the automation of construction through robotics, additive manufacturing and building information modelling.

The most sustained treatment of mobility is AI and the Mobility Revolution, which models urban transport as a set of decision problems: autonomous vehicles, adaptive traffic signals, Hyperloop and high-speed rail, congestion management, electric-vehicle charging, ride-sharing, drones, autonomous freight, smart roads and incident detection. Related volumes address green transport within urban infrastructure and reinforcement learning for traffic control within broader models of urban development, with data privacy, interoperability and scalability treated as preconditions for wider deployment.

Empathy in Algorithms and AI with Integrity argue that fairness, transparency and accountability should be specified as design requirements of public-facing AI, and the theme runs through the other volumes as well. This work examines algorithmic bias in policing, transport and public services, formulates fairness-constrained reinforcement learning, adversarial debiasing and participatory forms of explainability, and considers how the digital divide shapes who benefits from smart infrastructure.