The FNR is pleased to announce that 4 of 6 proposals have been retained for funding in the 2025-2 JUMP Call, an FNR commitment of 1,295 MEUR. The FNR’s JUMP programme aims to bridge the technical and funding gap between research-driven discoveries and their commercialisation/ utilisation, thereby enhancing the impact of Luxembourg’s research on the economy and society.
Please note, since 2026, this programme is called TREK/JUMP.
Funded projects
| PI | Host institution | Project title | Acronym | FNR committed |
| Juan Merlano Duncan | University of Luxembourg | Daedaletz | DAEDALETZ | €212,000.00 |
| Keywords | Digital Beamforming, phased arrays, satellite communications, FPGA. | |||
| Abstract | The global NGSO and High-Throughput Satellite (HTS) market is poised for significant expansion, driven by increasing demands for bandwidth and low-latency data transmission. In particular, the European segment is projected to reach $9.8 billion by 2029, growing at a compound annual growth rate (CAGR) of 14.1%. This rapid growth reflects not only market needs but also major advancements in satellite technology. Recent innovations, especially those targeting drastic reductions in power consumption, are transforming satellite payload design. By lowering the energy required for onboard processing, these advancements reduce the need for large solar panels, complex electronics, and heavy batteries. This leads to significantly lower launch costs, and the reduced mass and simplified manufacturing processes provide substantial economic benefits for large-scale satellite production. Such innovations have become essential enablers for scalable deployments and faster market adoption. Digital beamforming and phased arrays represent the highest level of flexibility in antenna and payload design. These systems can generate up to one beam per antenna element, enabling simultaneous multi-region coverage for communications and wide-area monitoring for sensing applications. Their ability to place nulls in specific directions also reduces interference and significantly improves the Signal-to-Interference-plus-Noise Ratio (SINR). Despite these advantages, digital beamforming faces two major challenges: the power consumption associated with each RF chain and the computational cost of complex beamforming algorithms. Addressing these limitations has been a core focus of the SIGCOM group, which has invested significant research effort in developing energy-efficient digital beamforming algorithms and exploring advanced low-power processor architectures. One of the most impactful advancements is the use of the 2D Fast Fourier Transform (2D FFT) for beamforming, successfully implemented and validated by SIGCOM. Power savings achieved with this approach exceed 90% compared to conventional digital beamforming methods, with even greater savings as the number of antennas and beams increases. These savings translate directly into major benefits for satellite platforms. DAEDALETZ aims to develop the next generation of highly efficient digital beamforming payloads using 2D FFT-based algorithms implemented on energy-efficient hardware. The proposal outlines the pathway toward implementing and commercializing a solution based on SIGCOM’s advanced firmware, providing a transformative approach to energy-efficient beamforming and cost-effective satellite manufacturing. In addition, DAEDALETZ is particularly well-suited for emerging Direct-to-Device (D2D) and Direct-to-Vehicle (D2V) satellite communication services. These services require highly dynamic beam steering, a large number of narrow and flexible spot beams, low latency, and extremely energy-efficient onboard processing. The scalability and low power consumption of the 2D FFT beamforming architecture make it an ideal match for these requirements. As the D2D and D2V markets expand, driven by satellite connectivity for smartphones, IoT devices, land vehicles, maritime platforms, and aviation, an energy-efficient, massively multi-beam digital beamforming solution becomes essential. DAEDALETZ directly addresses this need by enabling the generation and management of a very high number of beams at minimal computational cost, supporting seamless connectivity for large numbers of mobile and low-power terminals. Overall, DAEDALETZ provides a disruptive technological capability for future satellite infrastructures by combining unprecedented energy efficiency, multi-beam scalability, and native compatibility with next-generation D2D and D2V services. | |||
| PI | Host institution | Project title | Acronym | FNR committed |
| Sebastjan Glinsek | Luxembourg Institute of Science & Technology (LIST) | Data Exploration And Visualization Environment | Delve | €284,000.00 |
| Keywords | data management, agentic AI, research infrastructure, FAIR data, R&D workflows, laboratory informatics, multi-modal integration, institutional memory, LLM orchestration, scientific data integration | |||
| Abstract | R&D laboratories across life sciences, manufacturing, and research institutions face critical challenges in managing heterogeneous experimental data. Scientists work with diverse instrument outputs (CSV, TXT, DAT, multi-dimensional arrays) that create fragmented, incompatible datasets. Current solutions—ELNs, LIMS, and analysis tools—address isolated workflow segments but fail to provide end-to-end integration, leaving researchers with manual data wrangling, duplicated effort, and knowledge loss when team members transition. DELVE is a domain-agnostic data management platform combining cloud-native infrastructure, agentic AI orchestration, and retrieval-augmented generation to automate schema inference, data integration, and analysis from heterogeneous sources. The core innovation lies in LLM-powered AI agents that convert natural-language requests into validated schemas, SQL queries, and visualizations—enabling non-programmers to perform complex data operations without coding. Institutional knowledge accumulates through persistent vector embeddings of protocols, publications, and past analyses, ensuring research continuity across personnel changes. This 18-month JUMP spin-off project will advance DELVE from TRL 4 to TRL 7 through pilot deployments in research labs, RTOs, and industry R&D facilities. We will validate 50-70% reduction in data preparation time, demonstrate FAIR-compliant metadata generation, and establish commercial readiness for multi-tenant SaaS and enterprise deployment models. The project addresses growing regulatory pressure for research data transparency (FAIR principles, GxP compliance) while creating a sustainable technology venture positioned to serve European R&D infrastructure modernization initiatives. | |||
| PI | Host institution | Project title | Acronym | FNR committed |
| Djamila Aouada | University of Luxembourg | AI-powered Personalised Learning Platform For Inclusive And Engaging Education | Wingos | €300,000.00 |
| Keywords | AI in Education; EdTech; Personalized Learning; Adaptive Learning; Gamification; Accessibility; UX Design; Learning Analytics; primary school; secondary school; workforce training; Multilingual Education; Cognitive Science; Digital Pedagogy; Participatory Design; Inclusive Education. | |||
| Abstract | Wingos tackles a growing mismatch between highly diverse learner populations and one-size-fits-all teaching practices, a challenge that is especially acute in multilingual and heterogeneous systems such as Luxembourg. Teachers are expected to differentiate instruction, manage several languages, and track individual progress, yet they work with fragmented digital tools, limited time, and generic AI systems that ignore curricula, pedagogy, and regulatory constraints. Wingos is an AI-powered learning platform that helps teachers, learners, and institutions cope with the real complexity of multilingual and heterogeneous education systems. It transforms existing materials into curriculum-aligned learning units, organises them in an adaptive knowledge graph, delivers personalized and accessible learning journeys to students, and offers schools a compliant, data-informed way to introduce AI into everyday teaching. For teachers, Wingos provides a unified environment that replaces disconnected tools by combining content creation, classroom management, progress tracking, and a shared resource library. For learners, Wingos adjusts difficulty and pacing, provides scaffolding when needed, and supports multilingual understanding through explanations and translations grounded in teacher materials. The platform prototype has been concepted with expert educators to address the needs of real classrooms. Early evaluations indicate increased learner engagement, clearer visibility on progress for educators, and reduced monitoring workload. | |||
| PI | Host institution | Project title | Acronym | FNR committed |
| Beatriz García Santa Cruz | University of Luxembourg | Behavioural Remedy Applied Into Trauma-related Sleep Disturbances Through Personalised Application Of Digital Therapy | BRAINPATH | €499,000.00 |
| Keywords | Digital health, Personalised medicine, Human-AI interaction in healthcare | |||
| Abstract | Sleep disturbances (including nightmares, nocturnal distress, and trauma-related sleep disruption) affect millions of people and are especially common among individuals exposed to traumatic events. Although effective behavioural treatments exist, access is limited due to a shortage of specialised clinicians. BRAINPATH introduces a personalised digital therapeutic that delivers validated behavioural sleep interventions through an adaptive recommendation engine, trauma-informed emotional pacing, and interactive psychoeducation. The system adjusts content to each user’s symptom profile and sleep patterns, ensuring both relevance and clinical safety. Developed through a collaboration between University of Luxembourg, MIT, and the U.S. Veterans Affairs (US), the solution undergoes early expert validation and prepares the groundwork for future clinical integration. BRAINPATH aims to expand access to evidence-based care by offering a scalable, personalised, and clinically aligned digital intervention for people experiencing trauma-related sleep disturbances. | |||