Project overview

The rapid expansion of Electric Vehicle Charging Infrastructure (EVCI) introduces significant cyber security vulnerabilities, especially in the widely deployed Open Charge Point Protocol (OCPP). There is a shortage of realistic, protocol-specific datasets for training AI-powered intrusion detection systems.

Simultaneously, organisations need more efficient, AI-supported tools to assess and improve their cyber maturity in both critical and non-critical infrastructure environments. This internship addressed both challenges by translating academic research into practical industry applications at Templar International Group.

Activities

• Actively contributed as an embedded researcher within Templar International Group’s advisory and research team.

• OCPP / EVCI Security Project: Designed and implemented a novel hybrid (virtual + physical) testbed combining Virtual Machines, that represent Electric Vehicle Charging Stations (EVCS) and their Central Management Systems (CMS), real OCPP compliant charging units, CP-50 EVSE testers representing physical Electric Vehicles (EVs), and simulated EVs. Conducted realistic attack simulations (Address Resolution Protocol (ARP) spoofing, Man in The Middle (MiTM), replay, message injection, etc.) and generated a rich, labelled dataset with features like, semantic fields, FSM transitions, timing metrics and network characteristics. This work directly supports the development of context-aware AI intrusion detection models for EV charging infrastructure. 

• CMAD Automation Project: Led the conceptual design and development of a prototype for automating Templar’s Cyber Maturity Assessment and Diagnostics (CMAD) tool. This included AI-assisted workflows for evidence collection, scoring logic, automated reporting, standards mapping, and governance-aware AI integration suitable for handling sensitive client data.

• Participated in strategic discussions on AI integration in cyber security services, developed executive briefings and professional documentation, and supported international research dissemination.

• Translated doctoral-level research into commercially relevant concepts while maintaining strict confidentiality standards.

  
  

Impact

The internship delivered dual, mutually reinforcing streams of value:

1. OCPP-Centric Hybrid Testbed & Dataset: Produced a semantically rich dataset (now publicly available on GitHub) that captures both benign and stealthy adversarial OCPP traffic. It outperforms existing IoT/EV datasets in protocol semantics, FSM modelling, and realistic attack representation. This provides a strong foundation for training explainable AI-based IDS/IPS models capable of detecting sophisticated threats (MiTM, spoofing, protocol manipulation) in critical EVCI environments.

2. CMAD Automation Prototype: Developed an AI-enabled prototype that modernises Templar’s cyber maturity assessment process. It reduces manual effort, improves consistency and scalability, and supports better digital evidence handling while aligning with recognised security frameworks.

   
   

Dimensions of Impact:

• Organisational / Economic: Accelerated Templar’s CMAD product roadmap and enhanced their research-led consultancy offerings. The OCPP work strengthens Templar’s capability in cyber-physical systems security advisory.

• Sector / Policy: Improves resilience of EV charging infrastructure (a key part of sustainable energy transition) against cyber threats that could impact the power grid and public safety.

• Research / Knowledge: Bridges academia and industry by providing an open, reproducible dataset for the wider research community and demonstrating practical pathways for responsible AI use in cyber security.

• Social / Environmental: Contributes to safer, more trustworthy EV adoption and supports secure decarbonisation of transport.

The two projects reinforced each other: insights from EVCI threat modelling informed broader cyber maturity assessment thinking, while CMAD work highlighted the need for domain-specific AI applications.

Future work

• Continue iterative updates and expansion of the OCPP dataset on GitHub (adding OCPP 2.0.1 support, encrypted traffic analysis, and more attack scenarios).

• Further development and refinement of the CMAD automation prototype towards a production-ready tool at Templar International Group.

• Joint research and potential publications or pilots combining the dataset with AI maturity assessment frameworks. 

• Extended collaboration with Templar on cyber-physical security services and additional academic-industry projects. 

• Invite other researchers and organisations to use the public dataset for benchmarking IDS models and to collaborate on testbed enhancements.

Outcomes/outputs

• Publication: “A Novel OCPP-Centric Hybrid Testbed and Dataset for EV Charging Infrastructure Security Threats – Feasibility Testing” (IEEE).

• Dataset: Publicly released OCPP-Centric Hybrid Testbed Dataset – Al-Mhiqani/OCPP-Centric-Hybrid-Testbed-for-EV-Charging-Dataset- (ongoing updates).

• Prototype: AI-enabled CMAD Automation Prototype (currently in development at Templar International Group).

• Presentation of my research work, including the hybrid testbed, dataset generation, and CMAD automation prototype, to both academic faculty at ENSIAS (École Nationale Supérieure d’Informatique et d’Analyse des Systèmes) and industry partners in Rabat, Morocco. This engagement reinforced shared interests in applied cyber security research, critical infrastructure protection, AI integration in security, and fostered stronger international collaboration, knowledge exchange, and future joint projects.

• Strategic Deliverables: Executive briefings, technical architecture, and knowledge transfer materials on AI integration and cyber-physical security.

• Collaborations: Strengthened and expanded partnership between University of Huddersfield, Templar International Group, and the international academic partner ENSIAS

• Capability Building: Enhanced Templar’s internal expertise in research-informed product development and responsible AI adoption.