Research
How do independently-developed digital twins become an ecosystem that software agents can actually reason about? That question was the subject of my Ph.D. thesis, and I keep pulling on the thread.
I'm part of the Pervasive Software Laboratory (PSLab) at the Department of Computer Science and Engineering, University of Bologna.
Digital Twin Ecosystems
Digital twins are usually built one at a time, for one system, by one team. The moment you want two of them to cooperate, you discover that neither can describe itself to the other.
I work on hypermedia and Linked Data as the interoperability layer: self-describing digital twins that publish their own affordances, so a client — human or agent — can discover what a twin represents and what it can do at runtime rather than at integration time.
- Towards Linked Data for Ecosystems of Digital Twins (2024) — the modelling side of the problem.
- Hypermedia-based digital twin ecosystems: A proposal for digital twin interoperability (2026) — the full proposal, in Future Generation Computer Systems.
- HWoDT Framework: A toolchain to build interoperable Digital Twin Ecosystems (2025) — the toolchain that implements it.
- Digital Twin Continuum: a Key Enabler for Pervasive Cyber-Physical Environments (2024) — pushing twins across the edge-cloud continuum.
- Distributing intelligent functionalities in the Internet of Things with agents and Digital Twins (2025) — where the intelligence should sit, between agents and twins.
Engineering Multi-Agent Systems
The BDI model gives agents a vocabulary for beliefs, desires and intentions — but the frameworks that implement it have real engineering gaps.
- Concurrency — why we should control it, and what current frameworks actually do.
- Explainability — a multi-level framework for engineering and understanding BDI agents, extended to explanations between agents.
- Anticipatory reasoning — Reflexive anticipatory reasoning by BDI agents (2025), and what anticipation means in multi-agent contexts (Best Paper, WESAAC 2025).
- Languages — JaKtA, a BDI agent DSL embedded in pure Kotlin: the language and how it blends with OO and functional code.
Web of Things & Hypermedia MAS
If the environment of an agent is the Web, then hypermedia controls are its affordances. I'm interested in how agents discover and use them, and what it takes for people who aren't programmers to configure such systems.
- Domain-Expert Configuration of Hypermedia Multi-Agent Systems in Industrial Use Cases (2023) — putting industrial hypermedia MAS in the hands of domain experts.
- Towards Framing the Agents & Artifacts Conceptual Model at the Knowledge Level: First Ideas and Experiments (2023) — the Agents & Artifacts model, framed at the knowledge level.
- Towards Agents' Embodiment in Hypermedia Multi-agent Systems (2025) — what embodiment means in a hypermedia space.
- The Gap Between BDI Agents and Semantic Hypermedia and What We Can Do About It (2025) — and what we can do about it.
Applications
The abstractions get tested against real domains.
- Healthcare — operating-room management and oncologic pharmaceutical supply chains.
- Industry 4.0 — modular engineering of industrial twins and cyber-physical interaction engineering.
- Smart farming — Semantic Knowledge for Autonomous Smart Farming (2022).
- Urban digital twins — the socio-demographic extension of the Urban Digital Twin of Bologna, at the Alma Human AI Research Center.
Ph.D. thesis
Digital Twins Ecosystems for Intelligent Applications in Healthcare University of Bologna, 2025. Advisor: Prof. Alessandro Ricci.
Projects
- JaKtA — BDI agent-oriented programming as a Kotlin internal DSL
- HWoDT — toolchain for interoperable Digital Twin ecosystems
- Pervasive Software Laboratory — the group I work in