Project Genesis and EU Funding Context
The XRisis project emerged from a critical recognition that humanitarian emergency response training faced fundamental limitations that technology could address. Nuwa secured funding through the EU Horizon Europe CORTEX2 Open Call Track 2 cascade funding programme in July 2024, establishing a partnership with Action Contre la Faim's Emergency Readiness and Response Unit to develop an immersive simulation platform for crisis management training. The CORTEX2 consortium, led by the German Research Centre for Artificial Intelligence (DFKI), provided not only financial support but access to cutting-edge real-time cooperation technologies including Alcatel Lucent Enterprise's Rainbow CPaaS communication platform, DFKI's Video Call Alternative Appearance avatar system, CEA's Conversational Virtual Agent dialogue platform, and Linagora's meeting summarisation capabilities. This European Commission-backed initiative ran from August 2024 through June 2025, encompassing three distinct phases: requirements gathering and design, platform implementation with CORTEX2 integration, and comprehensive validation with emergency response professionals. The project represented a significant investment in demonstrating how extended reality and artificial intelligence could transform humanitarian training delivery, reducing costs whilst improving realism and accessibility for crisis response organisations globally. What distinguished XRisis from conventional training technology initiatives was its foundation in real operational requirements gathered directly from humanitarian field practitioners, ensuring that technical innovation served genuine user needs rather than pursuing technology for its own sake. The partnership structure positioned Nuwa as the technical development lead whilst Action Contre la Faim provided domain expertise, scenario development, and access to emergency roster personnel for validation exercises, creating a collaboration model that balanced innovation with practical applicability.
Humanitarian Training Gaps and Operational Context
Traditional humanitarian emergency response training operates under severe constraints that compromise preparedness effectiveness and organisational reach. Conventional simulation exercises require significant financial resources for venue hire, facilitator travel, participant accommodation, and scenario equipment, with costs often exceeding tens of thousands of euros for a single multi-day workshop serving fewer than twenty participants. Logistical complexities multiply when training involves geographically distributed teams across multiple countries, requiring coordination of schedules, visa arrangements, and travel logistics that can delay critical upskilling by months. Realism suffers because physical simulations cannot replicate the full complexity of crisis environments, the psychological pressure of decision-making under extreme conditions, or the coordination challenges of managing distributed response teams across unstable communication infrastructure. Action Contre la Faim's Emergency Readiness and Response Unit identified these limitations as barriers to achieving their strategic goal of ensuring 80% of country offices maintain up-to-date Emergency Preparedness and Response Plans, with conventional training delivery unable to scale sufficiently to meet growing demand across their global operations spanning 56 countries. The COVID-19 pandemic had demonstrated both the necessity of remote collaboration capabilities and the inadequacy of existing videoconferencing solutions for complex collaborative tasks requiring spatial awareness, document sharing, and multi-stakeholder coordination. Local staff in resource-constrained contexts faced particular disadvantages, lacking access to intensive face-to-face training opportunities available to headquarters personnel whilst bearing primary responsibility for implementing emergency responses in their communities. The humanitarian sector recognised that every dollar invested in disaster preparedness saves approximately seven dollars in response costs, yet training budgets remained insufficient to deliver preparedness at the scale and frequency required to maintain organisational readiness. These converging pressures created a compelling case for technology-enabled training delivery that could maintain pedagogical effectiveness whilst dramatically reducing cost barriers and logistical friction. Nuwa's proposal to address these gaps through cooperative real-time XR experiences aligned directly with Action Contre la Faim's strategic priorities for emergency preparedness innovation and capacity strengthening.
CORTEX2 Platform Architecture and Technical Integration
The XRisis platform architecture integrated four distinct CORTEX2 enabling technologies within a Unity-based multi-user environment running on European cloud infrastructure hosted in Germany and France. Alcatel Lucent Enterprise's Rainbow CPaaS formed the communication backbone, providing secure real-time voice and video connections across all user endpoints through integration with Rainbow's C# SDK, enabling seamless audio and video experiences whether participants joined via XR head-mounted displays, desktop applications, or web browsers. DFKI's Video Call Alternative Appearance technology enabled participants and facilitators to maintain presence and identity during training sessions without exposing personal video feeds, using MediaPipe for real-time face tracking and Ready Player Me avatars to create privacy-aware yet expressive communication channels that preserved the social dynamics essential for collaborative learning. CEA's Conversational Virtual Agent, integrated directly into Rainbow and appearing as a participant in multi-stakeholder calls, provided AI-based dialogue support that could recall core documentation, provide contextual guidance, and answer questions about emergency management procedures, effectively embedding institutional knowledge into the training environment through large language model capabilities grounded in Action Contre la Faim's operational documentation. Linagora's summarisation agent processed recordings of training sessions to generate automatic summaries in PDF format, supporting post-exercise analysis and learning reinforcement without requiring facilitators to manually document all participant interactions and decisions. The technical stack coordinated multiple components: a Unity server managing the multi-player engine across all endpoints, a WebSocket server distributing objects and state synchronisation between simulation environments, and a web application providing simulation coordination and inject management interfaces for facilitators controlling scenario progression. Infrastructure architecture decisions prioritised European data sovereignty, deploying on bare-metal cloud service providers within EU jurisdictions to maintain compliance with GDPR requirements and organisational data protection policies governing humanitarian operations. This integration represented genuine technical innovation, not merely assembling existing components but adapting research-stage CORTEX2 technologies for production deployment in a domain with strict reliability, security, and usability requirements that exceeded typical proof-of-concept demonstration standards.
Pilot 1: Arrival Briefing and Emergency Management Foundations
The first pilot phase introduced participants to fundamental emergency management concepts through individual interactions with an AI-powered virtual mentor. Each participant entered the Alaris headquarters virtual environment alone, receiving a group induction briefing from a facilitator through Rainbow CPaaS before engaging in one-to-one dialogue with Mentor Maud, an AI avatar powered by CEA's Conversational Virtual Agent platform. Mentor Maud explained key emergency concepts including the distinction between rapid-onset disasters, slow-onset crises, and protracted emergencies, walking participants through the emergency management cycle phases of preparation, alert, response, recovery, and mitigation. The virtual environment presented participants with interactive visualisations of different crisis scenarios, allowing them to explore emergency contexts ranging from sudden flooding events to gradual drought-induced food insecurity, building conceptual understanding through immersive representation rather than purely textual instruction. Learning objectives centred on knowledge recall: by session completion, participants needed to demonstrate ability to identify emergency management cycle phases, explain the purpose of Emergency Preparedness and Response Plans, and recognise the roles of different coordination mechanisms including Emergency Cells, Rapid Emergency Action Teams, and Multi-Sectorial Assessments. The AI dialogue system enabled participants to ask questions in natural language, with Mentor Maud drawing responses from Action Contre la Faim's emergency management documentation, Standard Operating Procedures, and training materials previously indexed and embedded as knowledge sources. Validation results from the May 2025 Paris workshop revealed that this pilot phase delivered limited incremental value compared to conventional e-learning delivery, with participants rating the informational briefing from an AI avatar at only 3.2 out of 5 for added training value, suggesting that immersive technology provided marginal benefits for theoretical knowledge transfer that could be accomplished equally well through existing digital learning platforms. Participant feedback consistently indicated that the elaborate virtual environment created unnecessary cognitive load when the primary task involved absorbing conceptual information, with several noting that the VR headset became uncomfortable during extended passive listening periods. This finding proved valuable for informing future development priorities: theoretical training delivered through conventional modalities frees immersive technology budgets for application areas where XR provides unique value that justifies the additional complexity and cost. The pilot successfully validated the technical integration of CEA's Conversational Virtual Agent with Rainbow communication infrastructure and demonstrated that AI dialogue capabilities could effectively deliver domain-specific knowledge, even though the immersive delivery channel itself did not enhance learning outcomes for this particular application.
Pilot 2: Collaborative Alert and Response Strategy Development
The second pilot positioned teams of four participants in a collaborative virtual environment tasked with analysing an evolving humanitarian emergency and developing a coordinated response strategy. Participants entered the Alaris headquarters coordination office environment simultaneously, each assuming specific roles (Country Director, Programme Head of Department, Support Head of Department, Programme Manager) that reflected actual organisational structures in Action Contre la Faim country offices. The scenario narrative began with storm-induced flooding across the fictional Soffeta province, with media reports indicating initial displacement of 8,000 people and government estimates projecting impact on up to 200,000 individuals, requiring the team to activate emergency protocols and position for rapid response. Participants received simulated emails from local partners describing infrastructure damage and contamination concerns, news articles reporting evolving casualty figures and displacement patterns, and video calls via Rainbow CPaaS from role-players representing the Country Director (physically absent from the country) and Emergency Readiness and Response Unit personnel in Paris, all of whom appeared through DFKI's Video Call Alternative Appearance system as avatars rather than direct video feeds. The team needed to establish an Emergency Cell, conduct initial information analysis, determine whether to launch a Multi-Sectorial Assessment, and outline a response strategy addressing water, sanitation, food security, and shelter requirements whilst coordinating with government authorities and other humanitarian actors. Training outcomes emphasised applied knowledge and collaborative decision-making under pressure: participants needed to demonstrate ability to analyse impending shocks, take appropriate action following organisational protocols, design response strategies meeting affected population needs and minimum humanitarian standards, and communicate effectively within team decision-making processes. The collaborative working element proved valuable but encountered implementation challenges, with participants noting that co-located physical presence in the same room whilst attempting to collaborate in a virtual environment created awkward dynamics, suggesting that future iterations should optimise either for fully remote participation or for in-person collaboration using desktop interfaces rather than attempting to bridge both modes simultaneously. Participants rated collaborative work in the VR coordination office at 3.6 out of 5 for added training value and the interactive response strategy tool at 3.4 out of 5, indicating moderate value that justified continued development whilst highlighting substantial room for interface refinement and workflow optimisation. The pilot successfully demonstrated Rainbow CPaaS's capability to support multi-modal communication across XR devices, desktop applications, and web browsers, with participants moving fluidly between one-to-one conversations, small group discussions, and full team coordination meetings. Technical integration challenges emerged around state synchronisation when participants manipulated shared documents and planning tools simultaneously, requiring WebSocket server optimisations to prevent conflicting updates and ensure all team members maintained consistent views of evolving response plans.
Pilot 3: Implementation Simulation and Stakeholder Negotiations
The third pilot immersed individual participants in response implementation scenarios requiring negotiation, problem-solving, and interpersonal skills with AI-powered stakeholders. Participants entered role-specific virtual environments representing field locations at day 14 of the emergency response, confronting challenges typical of the implementation phase including resource constraints, community coordination issues, supply chain complications, and inter-agency relationship management. Each participant received a task brief outlining a specific challenge (for example, negotiating access permissions with local authorities, coordinating distribution logistics with partner organisations, or addressing community concerns about response equity) and engaged in conversation with an AI avatar representing a key stakeholder powered by CEA's Conversational Virtual Agent platform. The AI avatars exhibited realistic conversational behaviours including cultural communication styles, bureaucratic caution, community advocacy perspectives, or logistical pragmatism depending on their role, requiring participants to adapt communication approaches, build rapport, manage expectations, and negotiate solutions acceptable to all parties. Learning objectives focused on soft skills that prove notoriously difficult to assess through conventional training methods: managing challenges during implementation, demonstrating cultural sensitivity, building stakeholder relationships under pressure, and making pragmatic decisions that balance ideal outcomes against operational constraints. This pilot phase received the highest validation ratings, with participants scoring practice of soft skills and negotiation with AI avatars at 4.2 out of 5 for added training value, substantially exceeding scores for the other pilot components and validating the hypothesis that immersive technology provides unique advantages for experiential learning that requires situated practice in realistic contexts. Participants reported that conversing with AI stakeholders felt sufficiently realistic to generate authentic emotional responses, including frustration with bureaucratic obstacles, satisfaction at finding common ground, and uncertainty about whether their negotiation approaches achieved optimal outcomes, creating pedagogical value through emotional engagement that theoretical case studies cannot replicate. Facilitators could observe participant interactions in real-time through Rainbow CPaaS video feeds and review automatically generated transcripts and summaries from Linagora's summarisation agent, enabling evidence-based debrief conversations about communication effectiveness, cultural awareness, and negotiation strategies without relying solely on participant self-reporting. The AI dialogue system sometimes struggled with non-native English speakers, particularly when participants exhibited regional accents or used domain-specific jargon, resulting in conversation breakdowns that required facilitator intervention to reset the scenario, highlighting the importance of continued refinement in speech recognition capabilities for international humanitarian contexts where multiple languages and linguistic backgrounds converge. Participants using desktop computers rather than VR headsets reported substantial value from the implementation pilot, confirming that the core benefits derived from realistic conversational AI rather than stereoscopic immersion, a finding that significantly improves deployment feasibility by eliminating the requirement for organisations to invest in expensive head-mounted display hardware.
Validation Methodology and Workshop Design
The validation approach combined quantitative metrics, structured feedback instruments, and qualitative debrief analysis to evaluate both technical integration and training effectiveness. The rehearsal workshop on 25 March 2025 in Paris involved the Action Contre la Faim project team in comprehensive testing of the latest software build, identifying interface issues, integration bugs, and workflow friction points that required resolution before final validation with emergency roster personnel. The final validation workshop on 14 May 2025 engaged eight participants drawn from Action Contre la Faim's internal emergency roster, organised as two teams of four people, representing roles aligned with Programme Lead, Logistics, and Finance functions typical of country office emergency response structures. Participant selection targeted individuals with relevant emergency deployment experience (five of eight participants had completed emergency deployments) and simulation exercise familiarity (six of eight had participated in three or more previous simulation exercises), ensuring the cohort possessed sufficient domain knowledge to provide informed assessment of training value rather than merely reacting to technology novelty. The workshop facilitated three distinct pilot experiences sequentially, with the XRisis team (Laurence Knoop providing training expertise, Guillaume Auvray managing technical operations, and Marie Guittet contributing Emergency Readiness and Response Unit perspective) observing participant behaviour, documenting technical issues, and supporting participants encountering difficulties. Evaluation instruments included the System Usability Scale, a validated ten-item questionnaire measuring perceived usability through statements like "I thought the system was easy to use" and "I found the system unnecessarily complex", with responses aggregated to produce scores ranging from 0 to 100. Added value assessment asked participants to rate five specific components (informational briefing from AI avatar, interactive response strategy tool, team collaboration in VR coordination office, soft skills practice with AI avatars, and facilitator debrief in VR environment) on a scale from 1 (no added value) to 5 (substantial added value) specifically regarding contribution to emergency management competency development. User satisfaction measurement asked for overall assessment of the VR simulation exercise concept using the same five-point scale, supplemented by open-ended questions inviting suggestions for improvement and identifying aspects that worked particularly well. Written surveys were supplemented with verbal group debriefs with participants and a separate debrief session with facilitators and the wider project team, enabling triangulation of feedback sources and capturing insights that might not emerge through structured questionnaires alone. This multi-method validation approach aligned with Action Contre la Faim's Monitoring, Evaluation, Accountability and Learning framework whilst satisfying CORTEX2 requirements for evidence-based assessment of technology integration effectiveness.
Quantified Results and Performance Analysis
The System Usability Scale assessment produced an average score of 59% across all participants, with six of ten users (including two facilitators) scoring 51% or higher and two users scoring 71% or higher, indicating usability fell below the 68% threshold generally considered acceptable for consumer applications but remained within the range typical for complex professional tools during initial deployment. This result validated concerns about interface complexity whilst demonstrating that motivated users with appropriate induction could achieve productive use despite usability friction, suggesting that refinement rather than fundamental redesign would likely yield acceptable user experience. Added value ratings showed substantial variation across pilot components, with the implementation simulation pilot (soft skills and negotiation practice with AI avatars) scoring 4.2 out of 5, collaborative team work in the VR coordination office scoring 3.6 out of 5, the interactive response strategy tool scoring 3.4 out of 5, facilitator debrief in VR scoring 3.3 out of 5, and informational briefing from AI avatar scoring only 3.2 out of 5, producing an overall average of 3.5 out of 5 (equivalent to 70% added value rating) across all components. This differentiation proved strategically valuable, clearly indicating that XR technology provides maximum training value for situated practice of complex interpersonal skills rather than knowledge transfer or structured planning activities, fundamentally informing future development priorities and market positioning. Overall user satisfaction with the VR simulation exercise concept averaged 3.3 out of 5 (66% satisfaction), with responses demonstrating high variance: 50% of users awarded the maximum rating of 5 out of 5 whilst others provided more modest assessments, suggesting the platform resonated strongly with certain user profiles whilst leaving others uncertain about value proposition relative to conventional training alternatives. Qualitative feedback reinforced quantitative patterns, with positive comments emphasising engagement value, novelty of experiencing realistic stakeholder interactions, and appreciation for safe practice opportunities without real-world consequences, whilst critical feedback focused on interface complexity, unclear task expectations during certain scenario phases, AI speech recognition failures with accented English, and unnecessarily large virtual environments creating cognitive distractions when tasks primarily involved conversation. Facilitators specifically noted that desktop user interfaces, whilst less immersive than VR headsets, proved entirely adequate for scenario coordination and might actually provide greater value for their workflow since they could more easily switch between scenario management, participant observation, and documentation tasks without removing head-mounted displays. These validation results substantially exceeded the performance levels typical of early-stage proof-of-concept demonstrations, achieving Technology Readiness Level 7 (system prototype demonstration in operational environment) rather than merely TRL 4-5 (laboratory validation), positioning the platform for commercial development with confidence that core value propositions had been validated through rigorous testing with target user populations.
Strategic Lessons and Development Recommendations
The validation process generated actionable insights that fundamentally shaped the platform evolution strategy and commercial development roadmap. The single most important lesson concerned selective application of immersive technology: not every training requirement benefits from XR delivery, and attempting to build a comprehensive platform covering all emergency management cycle phases would dilute resources across components with highly variable value propositions, whereas focusing investment on implementation simulation (the highest-rated component) creates defensible competitive advantage in an application area where XR provides unique capabilities. Participants and facilitators consistently emphasised that theoretical knowledge transfer, whilst necessary for competency development, delivers equivalent outcomes through conventional e-learning platforms without incurring XR development and deployment costs, recommending clear separation between theoretical training delivered through existing modalities and application training delivered through immersive simulation environments. Interface streamlining emerged as a critical success factor, with complexity cited as the primary usability barrier: future iterations needed to reduce visual clutter, simplify navigation mechanics, provide clearer signposting about task expectations at each scenario phase, and ensure participants could achieve productivity without extensive technical induction that consumed valuable training time. The finding that desktop computers provided adequate value for most training applications whilst VR headsets enhanced immersion without fundamentally changing learning outcomes profoundly influenced deployment strategy, confirming that the platform should support both modalities rather than requiring expensive hardware investments that many humanitarian organisations cannot justify within constrained budgets. Speech recognition quality represented a blocking issue for international deployment, with AI conversation failures frustrating participants and undermining scenario realism, requiring either substantial improvement in recognition accuracy for accented English or deployment of alternative interaction modalities (perhaps text chat interfaces supplementing voice) to ensure reliable operation across the linguistic diversity characteristic of global humanitarian operations. Participant feedback about collaborative dynamics when co-located individuals attempt to work together in virtual environments revealed important design considerations: the platform should optimise either for fully remote collaboration (where virtual presence provides clear benefits) or for co-located collaboration using desktop interfaces (where physical presence enables natural communication supplemented by digital tools) rather than attempting to bridge both modes in a single session. Environmental design guidelines emerged from feedback about unnecessarily detailed virtual spaces creating distraction: if a scenario primarily involves conversation with minimal spatial interaction, keep environments simple and focused; reserve elaborate 3D environments for scenarios requiring spatial navigation, object manipulation, or location-based decision-making where environmental complexity serves pedagogical purposes. Technical literacy concerns, whilst raised during design discussions, proved less significant than anticipated, with appropriate induction enabling participants across age and gender demographics to achieve productive use, suggesting that assumptions about technology adoption barriers sometimes reflect designer uncertainties rather than actual user capabilities. These lessons collectively reinforced the importance of user-centric design: regular exposure to actual users engaging with evolving prototypes generated insights impossible to derive from technical specifications or expert assessments alone, validating the investment in iterative design workshops even when they revealed uncomfortable truths about features that seemed promising in concept but delivered limited value in practice.
Platform Evolution: From XRisis to ImmErgenSim and SimExBuilder
The successful validation of XRisis as a proof-of-concept established the foundation for commercial platform development under the ImmErgenSim brand, which subsequently evolved into the SimExBuilder platform. Nuwa and Action Contre la Faim agreed to continued development beyond the CORTEX2 project conclusion, with a formal Service Level Agreement model targeting Q2 2026 for commercial launch, transitioning from research demonstration to sustainable service delivery. The SimExBuilder platform concept emerged from recognition that whilst XRisis validated humanitarian emergency training as a compelling initial use case, the underlying technology architecture could serve broader simulation exercise requirements across multiple sectors including healthcare emergency preparedness, industrial safety training, smart city planning, and educational scenario-based learning. SimExBuilder's value proposition centres on being a low-code simulation builder enabling organisations without extensive technical capabilities to design their own immersive training simulations through intuitive authoring tools, scenario templates, and configurable AI agent personalities, dramatically reducing the cost and technical skill barriers that currently limit simulation exercise adoption to well-resourced institutions. The platform architecture inherited XRisis technical foundations whilst expanding capabilities: continued integration with Alcatel Lucent Enterprise's Rainbow CPaaS for communication backbone, DFKI's Video Call Alternative Appearance for privacy-aware presence, CEA's Conversational Virtual Agent for AI dialogue, and Linagora's summarisation for automatic documentation, with Nuwa negotiating extended exploitation licences for these CORTEX2 technologies to support commercial deployment beyond initial research funding. Competitive advantage derives from humanitarian domain expertise, first-mover positioning in the emerging XR simulation exercise market, and the combination of desktop and VR headset compatibility that allows organisations to match deployment hardware to budget constraints rather than requiring universal expensive VR equipment. Market positioning targets CEOs, CFOs, and CIOs from global humanitarian emergency and response NGOs as primary decision-makers, with revenue models combining tiered organisational account subscriptions, professional services for custom scenario design, marketplace commissions for scenario template sales, and ongoing Service Level Agreements covering platform maintenance, content updates, and technical support. The platform achieved recognition through Unity's 2025 Unity for Humanity Grant, with Nuwa's XR Ireland sister brand receiving prestigious acknowledgement as one of ten winners from nine countries for ImmErgenSim's contribution to addressing global challenges aligned with UN Sustainable Development Goals, validating both the technical innovation and the social impact potential of the platform. This award provided not only financial support and marketing visibility but independent validation of the platform's readiness for scaled deployment, strengthening investor confidence and partnership development conversations. The evolution from CORTEX2-funded research through proof-of-concept validation to commercial platform development with major award recognition demonstrates successful execution of EU innovation funding objectives: public investment in early-stage research generating commercially viable solutions with sustained impact beyond initial grant periods.
Impact Pathway and EU Ecosystem Positioning
The XRisis project exemplifies the value of EU cascade funding mechanisms in bridging the gap between fundamental research and commercial application. The CORTEX2 Open Call structure provided essential resources for small-to-medium enterprises like Nuwa to access cutting-edge technologies developed by major research institutions including DFKI, CEA, and commercial innovators like Alcatel Lucent Enterprise and Linagora, creating collaboration opportunities that would be economically infeasible without public funding coordination. Regular mentoring sessions with CORTEX2 consortium leaders throughout the project duration provided technical guidance on integration approaches, troubleshooting support when encountering API limitations or performance bottlenecks, and connections to appropriate technical contacts within partner organisations, substantially accelerating development velocity compared to independent commercial development pathways. The project generated multiple forms of research output beyond the platform itself: peer-reviewed conference papers submitted to EuroXR 2025 and IDRC 2025 conferences contributing to academic knowledge about XR application in humanitarian contexts, Zenodo publications providing open-access documentation of evaluation methodologies and results for other researchers building on this work, and engagement with European XR professional networks through presentations at XR Expo Stuttgart, Viva Technology Paris, and Eirmersive meetups sharing practical implementation lessons with the broader immersive technology community. Dissemination activities reached diverse stakeholder audiences including policy makers interested in humanitarian innovation, public authorities evaluating training technology investments, humanitarian sector decision-makers assessing digital transformation opportunities, and corporate partners exploring XR deployment for their own training requirements, creating network effects that amplify impact beyond direct platform users. The project's contribution to establishing European leadership in socially impactful XR applications directly supports EU strategic objectives around digital sovereignty, technological innovation serving social good, and competitive positioning in emerging technology markets currently dominated by North American and Asian firms. Partnership deepening extends beyond the immediate project participants: relationships established through CORTEX2 consortium activities positioned Nuwa for potential involvement in follow-up Horizon Europe proposals addressing next-generation collaborative XR capabilities, cross-border XR service integration, and domain-specific application development for critical social sectors. The platform's architecture deliberately prioritised European data sovereignty by deploying exclusively on EU-based cloud infrastructure, maintaining GDPR compliance throughout, and selecting technology partners with European headquarters or strong European operations, aligning commercial strategy with regulatory environment and cultural preferences around data protection and digital rights. This positioning creates competitive advantage in markets where organisations face increasing scrutiny about data residency, processing transparency, and compliance with evolving digital regulations. The pathway from research funding through validation to Unity for Humanity recognition and commercial platform launch demonstrates the multiplier effect of strategic public investment: relatively modest CORTEX2 cascade funding (part of broader programme budget) catalysed development that attracted commercial partnership, international recognition, and sustainable business model development, creating long-term economic and social value substantially exceeding initial public expenditure.
Transformative Potential for Humanitarian Preparedness
The broader implications of XRisis extend beyond immediate training delivery improvements to potentially reshaping how humanitarian organisations approach emergency preparedness and capacity development. Current training delivery models concentrate opportunities in headquarters locations and regional hubs, creating geographic and institutional inequities where field staff in remote country offices receive substantially less training investment than international personnel, directly contradicting localisation commitments to strengthen in-country capacity and leadership. XR simulation platforms can democratise access to high-quality experiential training, enabling field staff to participate in realistic scenario exercises from their duty stations without requiring international travel, visa processing, or extended absence from operational responsibilities, shifting the capacity development paradigm from centralised delivery to distributed access. The platform's capability to present training in multiple languages and cultural contexts addresses a persistent challenge in multinational organisations where training materials developed for Western contexts often fail to resonate with staff operating in very different cultural and operational environments, requiring extensive adaptation that proves resource-intensive and often incomplete. AI-powered dialogue agents can be trained on organisation-specific documentation, standard operating procedures, and institutional knowledge in multiple languages, creating personalised learning experiences that reflect local operational contexts whilst maintaining alignment with global organisational standards and humanitarian principles. The cost reduction potential proves particularly significant for smaller organisations and national NGOs that cannot justify dedicated training departments and struggle to provide staff with emergency preparedness development opportunities available to larger international organisations, potentially levelling the playing field and improving overall sector capability to respond to crises with appropriately trained personnel. Scenario reusability transforms training economics: once designed and validated, simulation scenarios can be deployed repeatedly across multiple cohorts with marginal costs approaching zero, contrasting sharply with conventional in-person exercises where each iteration incurs full facilitator, venue, and logistics costs, fundamentally changing the cost-benefit calculation for intensive preparedness training. The platform enables progressive difficulty calibration, starting participants with foundational scenarios and gradually increasing complexity as competencies develop, supporting structured learning pathways rather than one-off workshop experiences, potentially enabling humanitarian organisations to implement career development frameworks with clear competency progression tied to demonstrated performance in standardised simulation exercises. Evaluation data from simulation performance creates unprecedented visibility into individual and organisational preparedness, enabling evidence-based deployment decisions, targeted remedial training for identified skill gaps, and aggregate analysis of systemic capability strengths and weaknesses across organisations and regions, supporting strategic workforce planning and capacity development investment allocation. Integration with existing Learning Management Systems and Monitoring, Evaluation, Accountability and Learning frameworks positions simulation training as a complement to rather than replacement for conventional approaches, embedding within existing institutional structures rather than requiring parallel processes. The transformative potential ultimately depends on adoption dynamics: will humanitarian organisations invest in new training modalities despite budget constraints and institutional conservatism? Will staff embrace immersive learning despite potential technology anxiety and preference for familiar formats? Early validation results suggest cautious optimism, with practitioners recognising clear value for specific applications whilst maintaining appropriate scepticism about overpromising technology's capacity to solve fundamentally human challenges around preparedness culture, organisational commitment, and individual motivation to invest in capability development for scenarios that may never occur.
Future Directions and Development Priorities
The immediate development priorities for SimExBuilder focus on addressing usability barriers, expanding scenario libraries, and establishing sustainable commercial operations. Interface streamlining will incorporate participant feedback about complexity, implementing progressive disclosure of advanced features, providing contextual help and guidance, redesigning navigation to reduce cognitive load, and ensuring new users can achieve basic productivity within fifteen minutes rather than requiring hour-long induction sessions that consume valuable training time. Scenario authoring tools represent critical capability for commercial viability, enabling clients to design custom training scenarios without requiring Nuwa's direct involvement in every scenario creation cycle, using template-based approaches, drag-and-drop environment assembly, visual scripting for scenario logic, and no-code AI agent personality configuration that democratises simulation design across organisations without dedicated technical staff. The platform will expand hardware compatibility beyond current desktop and VR headset support to encompass mobile devices for basic participation modes, augmented reality devices for mixed-reality training applications, and accessibility accommodations including screen reader support, alternative input modalities, and adjustable sensory intensity settings ensuring inclusive access regardless of physical capabilities. Integration expansion will pursue connections with organisational systems including Single Sign-On authentication providers, existing Learning Management Systems for training record integration, collaboration platforms like Microsoft Teams and Slack for seamless communication workflow, and data analytics tools enabling organisational leaders to track training completion, performance trends, and competency development across their workforce. The platform architecture will support multi-organisational and inter-agency simulation exercises, recognising that complex humanitarian emergencies require coordination across multiple organisations and UN cluster systems, enabling scenarios where participants from different organisations must negotiate, coordinate, and make collective decisions whilst maintaining appropriate information boundaries and organisational identities. Commercial go-to-market strategy will target initial deployment with Action Contre la Faim's global country office network, using their organisational footprint as reference implementation demonstrating value at scale before approaching other humanitarian organisations with validated case studies and performance evidence. Partnership development will pursue collaboration with humanitarian training consortiums including UN agencies' training arms, regional humanitarian networks, and professional development providers seeking to enhance their service offerings with immersive capabilities without developing competing platforms. Technical roadmap priorities include reducing system requirements to enable participation from locations with limited internet bandwidth (3G network support), improving AI speech recognition for non-native speakers and multilingual contexts, expanding language support beyond English and French to include Spanish, Arabic, and other languages prevalent in humanitarian operations, and developing offline-capable modes allowing scenario pre-loading for use in environments with intermittent connectivity. Research collaborations will continue engagement with academic institutions studying immersive learning effectiveness, contributing to evidence base about XR training outcomes, publishing findings in peer-reviewed venues, and maintaining connection to frontier research ensuring SimExBuilder incorporates emerging capabilities as they mature from laboratory demonstrations to production-ready technologies. The development roadmap remains grounded in user needs rather than technology possibilities, maintaining the user-centric design approach that proved essential during XRisis validation and resisting temptation to add features that demonstrate technical sophistication without delivering proportional user value.
Conclusion: From European Research to Global Humanitarian Impact
The XRisis journey from initial CORTEX2 funding through rigorous validation to Unity for Humanity recognition and commercial platform development demonstrates the complete research-to-impact pathway that EU innovation programmes aim to catalyse. What began as a technically focused proposal to integrate cooperative real-time XR technologies evolved through user-centric design into a platform addressing genuine operational needs in humanitarian emergency preparedness, guided by continuous engagement with domain experts who ensured innovation served practical requirements rather than pursuing technology for its own sake. The validation evidence provides confidence that immersive simulation training delivers measurable value for specific applications, particularly implementation phase soft skills that prove difficult to develop through conventional training methods, whilst also generating clear-eyed recognition that XR is not a universal solution applicable to all training requirements and organisations must thoughtfully determine where immersive delivery justifies additional complexity and cost. Partnership between Nuwa's technical capabilities and Action Contre la Faim's humanitarian expertise created synergies neither organisation could achieve independently, exemplifying the collaboration models that European funding mechanisms intentionally cultivate to accelerate innovation whilst grounding development in real-world applicability. The integration of CORTEX2 enabling technologies from multiple research and commercial partners demonstrated both the technical feasibility of assembling complex capabilities from distributed sources and the practical challenges of managing dependencies, reconciling architectural assumptions, and troubleshooting integration issues across organisational boundaries, providing valuable lessons for future collaborative technology development initiatives. Recognition through Unity for Humanity validates the platform's social impact potential whilst providing resources and visibility that accelerate commercial development and market positioning, creating momentum that transforms proof-of-concept into sustainable enterprise capable of delivering ongoing value beyond initial research funding. The platform's evolution toward SimExBuilder demonstrates how successful EU research investments can generate commercial outcomes with sustained impact, creating viable businesses that continue innovation, employment, and capability development long after public funding concludes. The humanitarian sector stands at an inflection point where digital transformation offers genuine potential to improve emergency preparedness effectiveness, reduce training delivery costs, democratise access to high-quality learning experiences, and ultimately enhance the capability of the global humanitarian system to respond to crises that affect millions of vulnerable people each year, making research investments like CORTEX2 and projects like XRisis not merely interesting technical demonstrations but essential contributions to strengthening the institutional foundations of humanitarian response in an era of escalating climate-driven disasters and geopolitical instability. The journey from research concept through validation to commercial deployment and international recognition represents exactly the impact pathway that European innovation programmes seek to enable, demonstrating that strategic public investment in early-stage technology development can catalyse sustainable enterprises delivering measurable social value whilst strengthening Europe's competitive position in emerging technology markets.
Related
Partners
Industries
Products
Technologies
Related Portfolio Projects
