QPIC-AI will implement its activities within the following Work Packages:
Work Package 1 : Project Management and Coordination
This WP ensures the overall management, administration, coordination, and execution of QPIC-AI.
It guarantees that the project progresses according to the Grant Agreement (GA), that results are achieved in line with the call objectives, and that collaboration between EU and Korean partners is well coordinated.
The WP also ensures risk management, quality assurance, ethical compliance, and FAIR data management across all activities.
Work Package 2 : PIC Hardware Design and Fabrication
The main objectives of WP2 are:
- To design, fabricate, and demonstrate a PIC based on a TFLN platform while relying on external laser sources and detectors.
- To integrate DPC and other essential photonic building blocks into a compact, fibre-coupled module.
- To provide the core hardware building block for phase-encoded QKD in fibre-optic networks, enabling stable operation and interoperability with FPGA-based control (WP3) and AI-assisted stabilisation modules (WP4).
Work Package 3 : FPGA-Based QKD System Development Prototype ready for integration/field use
- Design and Implementation: Develop an FPGA-based decoy-state, phase-encoded BB84 system, interfaced to the WP2 PIC front end (AMZI, DPC, TEC) and InGaAs single-photon detectors.
- Closed-loop Stabilisation (AI-assisted): Provide hooks for ML-assisted control (developed in WP4) to manage interferometer phase, polarisation and temperature, targeting V₀ ≥ 0.98, σT ≤ 30 mK and auto-relock ≤ 1 s. ML integration will occur in Year 3 within WP5 at SKT, but the FPGA will be designed from the start with this compatibility in mind
- Lab Performance Demonstration: Achieve continuous key generation over 25–50 km fibre at frep = 125 MHz, with raw sifted ≥ 30 kbps @ 50 km and QBER < 3% under stable conditions.
- Interoperability & Field Validation: Generate rich telemetry/log datasets (QBER drift, register maps, calibration cycles) for EU partners in Years 1–2. End-to-end field validation will be carried out in Year 3 under WP5.
- Prototype Delivery: Provide a bench-top prototype with GUI-based monitoring/diagnostics, with baseline functionality delivered by M18–24, prepared for AI/KMS integration in Year 3 and field use in SKT’s on-net trial (WP5).
Work Package 4 : AI-Assisted Drift Prediction and Control
The objectives of WP4 are:
- To develop lightweight, real-time AI algorithms that stabilise QKD performance against drift in phase, polarisation, and temperature.
- To embed these algorithms directly into FPGA/SoC hardware using ML-on-FPGA toolchains, ensuring low latency, portability, and explainability.
- To implement hybrid control strategies combining classical feedback (PID loops) and adaptive ML routines, achieving long-term autonomous operation.
- To validate AI models first in proxy/simulation environments (Years 1–2) and then as embedded hardware functions during full integration at SKT in Year 3.
- To provide datasets and metrics supporting field validation (WP5) and inputs to standards/certification (WP6).
Work Package 5 : Integration, Testing and Validation
- Definition of the training strategy and methodology of HellasQCI
- Design of the training stages and implementation rounds for HellasQCI trainers groups
- Implementation of the training program on QKD principles for the research and academic communities
- Implementation of the training program on Quantum Security for security and IT experts
- Implementation of the training program on QKD applications for end users and public authorities
Work Package 6 : Standards, EU–KR Collaboration, Dissemination and Exploitation
- To ensure that QPIC-AI outcomes are transferred into international standardisation efforts (ETSI, ITU-T, ISO/IEC, CEN-CENELEC) and certification frameworks.
- To strengthen EU–ROK collaboration through joint testbeds, shared roadmaps structured researcher exchanges and training activities.
- To promote knowledge transfer via stakeholder workshops and open reference designs.
- To maximise project visibility, communication, and dissemination through coordinated outreach at EU and KR levels.
- To define a clear exploitation and IPR strategy, ensuring sustainability and industrial uptake beyond the project lifetime.