Fourteen years of the AAL Programme have created one of Europe’s richest evidence bases on what drives—or derails—innovation in health and care for older adults. While technologies have evolved, the core lessons about adoption, regulation, and long-term impact remain remarkably consistent. This article distills those hard-earned insights into practical, transferable guidance for future initiatives.
Whether you’re designing a product, managing a funding programme, or shaping systemic policy, the following sections unpack what worked, what didn’t, and what to do differently next time:
The lessons that follow distil this experience into practical guidance for newcomers (or not) — whether tech developers, care investors, or ecosystem orchestrators — based on what consistently worked, and what didn’t.
The most successful teams engaged all three user groups—older adults, carers (formal and informal), and payers—from the very beginning. Projects that involved these actors continuously, starting in month one, were 2.1× more likely to reach market and achieved 30–40% higher sustained use. This means co-design isn’t a workshop at the end—it’s an ongoing dialogue that shapes every phase, from concept to post-launch.
Equally important is recognising that older adults are not a monolithic group. Usability needs vary drastically across physical, cognitive, and digital literacy levels. Solutions that ignored frail or low-literacy users during development often faced late-stage usability shocks and were forced into costly redesigns—or never made it past pilot.
Innovation in this sector succeeds when it fits into existing care and data ecosystems. That means developing modular, standards-based systems from the outset. Projects using interoperable protocols like FHIR®, MQTT, and open APIs reached commercial deployment 64% of the time, versus 38% for closed, monolithic systems. With the European Health Data Space becoming operational in 2025, certified interoperability will no longer be optional—it will be a legal requirement for data sharing.
Closed or proprietary solutions already face resistance from hospitals and insurers, and soon may be excluded altogether from secondary-data markets. Teams that built with flexibility and openness baked in were better positioned to scale and adapt.
Too many promising projects were delayed or derailed by regulatory issues left until too late. Under the EU AI Act (Reg. 2024/1689), most monitoring or decision-support systems targeting older people are classified as “high-risk.” That comes with specific obligations: bias testing, compliance documentation, and EU database registration—all before market entry.
Fixing these gaps post-prototype caused 6–12 month delays in a third of projects. In more severe cases, ethics boards froze recruitment or reclassified the product as a medical device, adding 12–18 months to timelines. Teams that embedded privacy-by-design and completed AI/GDPR documentation from the start avoided these setbacks.
Living-lab pilots conducted in real homes or care settings for at least six months consistently produced user retention over 60%. In contrast, pilots limited to controlled environments often failed to exceed 40% retention. Realistic testing environments not only validate usability—they build trust with institutional buyers and support reimbursement discussions.
Still, testing needs to be meaningful. Projects that relied heavily on online surveys without deeper engagement contributed 70% of all user counts in some cohorts, yet contributed little to understanding real-world outcomes such as quality of life improvements, reduced care burden, or time savings for professionals. Real-world use matters more than inflated survey figures.
Solutions that reached the market typically included a demand-side partner—an insurer, hospital, or municipality—who helped shape the business model. These projects closed follow-on funding 30% more often, and transitioned from pilot to contract twice as fast as others. Payers bring not just money, but legitimacy and insights into procurement processes.
Conversely, relying on older adults to pay out of pocket proved unsustainable. Consumer-only models usually plateaued at under 3,000 users and rarely lasted beyond initial funding. The sector functions through institutional purchasing, not consumer apps.
Solutions with inclusive interfaces—voice controls, large-font screens, TV compatibility—and those that budgeted for digital-skills coaching had far higher adoption, especially among users with limited e-health literacy. Studies showed that these users were 2–3× less likely to adopt standard apps, but their engagement could be restored through design adaptation or caregiver dashboards.
Ignoring localisation also cost projects dearly. Teams that skipped language or cultural adaptation lost up to two-thirds of non-native users, and fixing the issue after launch proved more expensive than building in multilingual capability from the start.
Hardware risks are real. Projects that depended on single suppliers or specific cloud APIs saw entire value propositions collapse when those vendors exited the market. In contrast, teams that used commodity sensors and dual-supplier strategies kept running smoothly.
Similarly, projects that exceeded 10,000 users had one thing in common: hybrid financing models. These typically included a mix of public contracts, insurer support, and optional family subscriptions. Such models reflect how long-term care is financed in the EU and help projects survive market volatility.
No team succeeds alone. Regions that had ecosystem orchestrators or strong living-lab networks helped projects scale cross-border 30% faster and attracted follow-up ERDF funding. These environments offer integration pathways, not just visibility.
At the same time, access to tailored business support also mattered. Teams that participated in programmes like AAL2Business or EIC Market-Adoption Accelerator were 2.3× more likely to raise private capital within a year. These pipelines help convert prototypes into investable propositions.
Perhaps the most overlooked—but damaging—failure was the lack of post-grant maintenance. Projects without a clear support structure or service contract disappeared from use within 18 months, squandering earlier progress and evidence. Sustained impact requires a realistic operational model—not just a good prototype.
The AAL Programme legacy is a rich mix of institutional good practice, proven support tools and persistent structural bottlenecks. Future R&I and institutional-change partnerships—especially Horizon Europe’s “Transforming Health & Care Systems” and forthcoming “Demographic Change” initiatives— can build on nine transferable success factors and should avoid seven persistent challenges. Each persistent challenge is accompanied by concrete lessons drawn from the programme’s experience, offering practical guidance for future initiatives aiming to improve their effectiveness, sustainability, and impact.
Clear, adaptive governance enabled the programme to evolve from a fragile administrative start-up into a functioning strategic platform. Small, iterative changes—such as refining roles across the General Assembly, Executive Board and Management Unit, and centralising operations through Salesforce—cut time-to-contract from over a year to under six months in the fastest instruments. Dashboards and compliance audits also improved transparency and performance.
Challenge-led, co-creative calls ensured that funding responded to real-world needs. Calls framed around ageing-related challenges instead of technology classes invited applicants to co-design solutions with users, caregivers and payers. This significantly improved the usability and adoption of project outcomes.
A dedicated business-acceleration pipeline, AAL2Business, offered coaching tailored to the maturity level of each project. Lean Startup Academies helped projects test and refine how well their solutions addressed real user needs, and the Go-to-Market Launchpad prepared projects for investors. Projects that followed this path were over twice as likely to attract private follow-on funding.
Pan-European ecosystem events, like the AAL Forum (later EWAHA), built lasting relationships. These gatherings created a shared vocabulary and generated consortia that remained active for years. Trust, visibility, and alignment across care ecosystems improved noticeably.
The Monitoring–Evaluation–Impact (MEI) framework helped transform programme steering. Instead of measuring only inputs and outputs, MEI dashboards tracked real-world results like product launches and cost savings. These data loops supported call redesign and strategic corrections.
AAL’s early investment in digitalisation provided structural resilience. Because evaluation, contracting and monitoring were already cloud-based, the programme absorbed the COVID-19 shock without delays and actually increased its outreach at lower cost.
The requirement for trans-national consortia with strong SME involvement (≈40 %) diversified the R&I base. More than 60 % of the 293 funded SMEs had not participated in Horizon 2020 before.
Small Collaborative Projects (SCPs) offered a nimble option: shorter timelines, smaller budgets, faster cash flow, and the chance to scale promising results into full-size projects.
Embedding early ethics and data privacy guidance reassured both evaluators and public buyers. No GDPR-related stoppages occurred, and trust-by-design approaches accelerated uptake in sensitive environments.
The use of a “virtual common pot” with divergent national rules led to chronic inefficiencies. In slower countries, up to 40 % of budgets went unspent, while fast-contracting countries dominated. This discouraged balanced consortia and reinforced geographic inequalities.
→ Future programmes should set binding time-to-contract limits across countries and allow reallocation of unused funds to maintain programme-wide performance and participation.
The path from pilot to scale remained long and fragmented. Fewer than a quarter of marketed solutions generated substantial revenue, and many spin-out SMEs closed within three years due to limited access to scale-up capital and fragmented reimbursement regimes.
→ Combine short, low-TRL (Technological Readiness Level) exploration vouchers (like SCP), followed by development grants focused on co-creation, and only then larger implementation grants once market-validation gates are met.
→ Linking later-stage funding to proven readiness, and combining grants with outcome-based or repayable finance, would help viable projects cross the final threshold to adoption.
Low demand-side participation created a mismatch between what was developed and what care providers or insurers were willing to buy. Many projects lacked payer insight or were over-engineered for real-world needs.
→ Future calls should require the presence of payers or procurers in the consortium and reward concrete financial commitments or co-ownership of exploitation plans.
Market intelligence and data tools like the Market Observatory arrived late and were under-used at first. Without a shared taxonomy or automatic data feeds, many project teams lacked clear figures for market size or competitor mapping.
→ To close this gap, a dedicated support budget should be secured from the outset to fund sector-wide intelligence, impact monitoring, and horizontal support actions immune to annual reallocations.
The Central Management Unit’s strategic bandwidth was limited. As its mandate grew, fixed staffing levels constrained long-term foresight and risked losing coherence between programme goals and operations.
→ Finance lightweight integrator teams (1–2 FTE) in each willing region to connect AAL-type projects with local testbeds, clusters and Living Labs.
Follow-up after project closure was fragmented. Only half of finished projects completed post-grant surveys, leaving large gaps in knowledge about long-term social or economic outcomes.
→ Programmes should budget for structured follow-up years in advance and automate impact tracking—e.g. through short annual dashboards—for several years after grant closure.
The transition to Horizon Europe created turbulence. Stakeholders faced delays in information, which led to pauses in investment or programme engagement across ecosystems.
Extra recommendations:
→ Clear legacy planning should begin early, defining what tools, data and roles will be handed over, and to whom, at least a year before the end of the programme.
→ Use standardised templates for freedom-to-operate, MDR (Medical Device Regulation) classification, reimbursement codes; link payments to completion.
Reflections on how thematic focus and cross-national delivery models affect inclusiveness, scalability, and systemic alignment.
Targeting funding specifically at older adults allows innovation to be guided by clearly identified needs and lived realities. Programmes like AAL demonstrated how this focus enables the development of tailored solutions—such as large-font interfaces and voice-controlled devices—that are tested under conditions like frailty or low digital literacy. User involvement increased significantly, leading to products that reflect actual aspirations for autonomy and connection. Dedicated schemes also proved commercially effective, with over half of funded prototypes reaching market stage. Moreover, these programmes generated broader system-level impacts, such as serving as testbeds for remote consent, data protection standards, and AI-enabled care, which continue to support EU-level initiatives and help SMEs access trial infrastructure.
However, this thematic focus can also reinforce segmentation, encouraging design approaches that are acceptable only within age-specific contexts and may hinder wider adoption. Solutions often remain locked in fragile business models; many failed to secure long-term payers and remained grant-dependent. Despite their inclusive ambition, such programmes frequently under-represent the frailest, rural, or low-income populations, especially those excluded by digital barriers. Regulatory classification as high-risk technologies further burdens small consortia with costly certification processes once funding ends. Finally, focusing exclusively on ageing may draw resources away from integrated, life-course strategies and fragment broader care systems.
Multinational programmes offer access to a wider range of skills, perspectives, and infrastructures than national schemes can provide alone. The AAL initiative, with partners from 23 countries, combined expertise in technology, healthcare, and business innovation, enabling cross-country pilots that tested solutions in diverse care settings. These collaborations help reduce R&D fragmentation, align national priorities, and accelerate responses to shared challenges. They also carry weight in shaping common standards; for example, AAL’s data-handling templates were quickly adopted by national health authorities, illustrating how joint programmes can set benchmarks that support harmonisation across the EU.
Yet, running a funding scheme across borders introduces significant administrative complexity. Time-to-contract can be long and uneven across participating countries, discouraging SMEs and leaving budgets underused in slower systems. Participation often skews towards countries with greater capacity, leaving others marginalised despite the potential benefits. Diverse legal and regulatory frameworks—particularly around data and medical devices—can create serious delays or even force the cancellation of pilot sites. Coordination costs also rise, and if governance is unclear, the administrative load fails to deliver value. Only with strong procedures and binding commitments can multinational efforts overcome these structural imbalances and operate efficiently.