The global shift to electric mobility is placing immense pressure on charging infrastructure, making sophisticated software architecture a critical enabler for reliable and scalable operations. However, current centralized software architectures struggle to balance the need for global orchestration with the need for local execution, leading to systems that are fragile during network outages and difficult to evolve. A critical research gap exists in the lack of a comprehensive architectural framework that systematically integrates trustworthiness with specific adaptability mechanisms required to handle volatile market drivers, such as evolving communication protocols, emerging Vehicle-to-Grid (V2G) services, and dynamic grid integration, without requiring fundamental system redesign. This doctoral thesis addresses this gap by answering the question: How can charging station management software be improved to enhance resilience and adaptability? The answer is a holistic, multi-faceted solution centered on a novel distributed cloud-edge-IoT architectural framework. The proposed framework strategically distributes intelligence across IoT, edge, and cloud layers. This design empowers individual stations with local autonomy for service continuity during network outages while providing Charge Point Operators (CPOs) with robust centralized management. Built on containerization and microservices, the framework's resilience is validated using formal methods, translating the architecture into a Component & Connector (C&C) model and employing a MaxSAT solver to identify and rectify critical dependencies. This research further contributes a structured catalog of fault-tolerance tactics (Restart, Replacement, Checkpoint, Replication, and Migration) mapped to the architecture's layers. Looking beyond simple fault recovery, this work proposes an evolved MAPE-K model for true component-level self-healing, paving the way for adaptive and antifragile systems. The practical applicability of the architecture is validated through scenario-based analyses, grounded in ethnographic studies with a leading CPO, which use sequence diagrams to model the system's effective mitigation of common operational failures. Finally, a detailed risk assessment of a real-world infrastructure identifies critical security vulnerabilities and provides an actionable roadmap for their mitigation, underscoring that trustworthiness is incomplete without robust security. Ultimately, this thesis delivers a validated architectural blueprint, a catalog of resilience patterns, a forward-looking self-healing model, and a practical risk management strategy. It provides a comprehensive answer to enhancing the resilience and adaptability of EV charging software, offering a robust foundation for the future of electric mobility.
La transizione globale verso la mobilità elettrica sta esercitando una pressione immensa sulle infrastrutture di ricarica, rendendo l'architettura software sofisticata un fattore abilitante critico per operazioni affidabili e scalabili. Tuttavia, le attuali architetture software centralizzate faticano a bilanciare la necessità di un'orchestrazione globale con l'esigenza di un'esecuzione locale, portando a sistemi fragili durante le interruzioni di rete e difficili da far evolvere. Esiste una lacuna critica nella ricerca dovuta alla mancanza di un quadro architettonico completo che integri sistematicamente l'affidabilità con specifici meccanismi di adattabilità necessari per gestire i volatili fattori di mercato — come l'evoluzione dei protocolli di comunicazione, i servizi emergenti di Vehicle-to-Grid (V2G) e l'integrazione dinamica della rete — senza richiedere una riprogettazione fondamentale del sistema. Questa tesi di dottorato affronta tale lacuna rispondendo al quesito: come può essere migliorato il software di gestione delle stazioni di ricarica per potenziarne la resilienza e l'adattabilità? La risposta è una soluzione olistica e poliedrica incentrata su un innovativo framework architettonico distribuito cloud-edge-IoT. Il framework proposto distribuisce strategicamente l'intelligenza attraverso i livelli IoT, edge e cloud. Questo design conferisce alle singole stazioni un'autonomia locale per la continuità del servizio durante le interruzioni di rete, fornendo al contempo ai Charge Point Operators (CPO) una gestione centralizzata robusta. Basato su containerizzazione e microservizi, la resilienza del framework è validata utilizzando metodi formali, traducendo l'architettura in un modello Component & Connector (C&C) e impiegando un risolutore MaxSAT per identificare e rettificare le dipendenze critiche. Questa ricerca contribuisce ulteriormente con un catalogo strutturato di tattiche di tolleranza ai guasti (Restart, Replacement, Checkpoint, Replication e Migration) mappate sui livelli dell'architettura. Guardando oltre il semplice ripristino dei guasti, il lavoro propone un modello MAPE-K evoluto per un vero self-healing a livello di componenti, aprendo la strada a sistemi adattivi e antifragili. L'applicabilità pratica dell'architettura è convalidata attraverso analisi basate su scenari, fondate su studi etnografici presso un CPO leader del settore, che utilizzano diagrammi di sequenza per modellare l'efficace mitigazione dei comuni guasti operativi da parte del sistema. Infine, una valutazione dettagliata dei rischi di un'infrastruttura reale identifica le vulnerabilità critiche di sicurezza e fornisce una roadmap operativa per la loro mitigazione, sottolineando come l'affidabilità sia incompleta senza una sicurezza robusta. In definitiva, questa tesi fornisce un modello architettonico validato, un catalogo di pattern di resilienza, un modello di self-healing lungimirante e una strategia pratica di gestione del rischio. Essa offre una risposta completa per migliorare la resilienza e l'adattabilità del software di ricarica dei veicoli elettrici, ponendo fondamenta solide per il futuro della mobilità elettrica.
Reasoning on architectural issues for electric vehicle charging stations of the future
Dini, Vick Pierce
2025/2026
Abstract
The global shift to electric mobility is placing immense pressure on charging infrastructure, making sophisticated software architecture a critical enabler for reliable and scalable operations. However, current centralized software architectures struggle to balance the need for global orchestration with the need for local execution, leading to systems that are fragile during network outages and difficult to evolve. A critical research gap exists in the lack of a comprehensive architectural framework that systematically integrates trustworthiness with specific adaptability mechanisms required to handle volatile market drivers, such as evolving communication protocols, emerging Vehicle-to-Grid (V2G) services, and dynamic grid integration, without requiring fundamental system redesign. This doctoral thesis addresses this gap by answering the question: How can charging station management software be improved to enhance resilience and adaptability? The answer is a holistic, multi-faceted solution centered on a novel distributed cloud-edge-IoT architectural framework. The proposed framework strategically distributes intelligence across IoT, edge, and cloud layers. This design empowers individual stations with local autonomy for service continuity during network outages while providing Charge Point Operators (CPOs) with robust centralized management. Built on containerization and microservices, the framework's resilience is validated using formal methods, translating the architecture into a Component & Connector (C&C) model and employing a MaxSAT solver to identify and rectify critical dependencies. This research further contributes a structured catalog of fault-tolerance tactics (Restart, Replacement, Checkpoint, Replication, and Migration) mapped to the architecture's layers. Looking beyond simple fault recovery, this work proposes an evolved MAPE-K model for true component-level self-healing, paving the way for adaptive and antifragile systems. The practical applicability of the architecture is validated through scenario-based analyses, grounded in ethnographic studies with a leading CPO, which use sequence diagrams to model the system's effective mitigation of common operational failures. Finally, a detailed risk assessment of a real-world infrastructure identifies critical security vulnerabilities and provides an actionable roadmap for their mitigation, underscoring that trustworthiness is incomplete without robust security. Ultimately, this thesis delivers a validated architectural blueprint, a catalog of resilience patterns, a forward-looking self-healing model, and a practical risk management strategy. It provides a comprehensive answer to enhancing the resilience and adaptability of EV charging software, offering a robust foundation for the future of electric mobility.| File | Dimensione | Formato | |
|---|---|---|---|
|
Reasoning on Architectural Issues for Electric Vehicle Charging Stations of the Future.pdf
accessibile in internet per tutti
Descrizione: Reasoning on Architectural Issues for Electric Vehicle Charging Stations of the Future
Dimensione
5.51 MB
Formato
Adobe PDF
|
5.51 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/254497