Over the past three years, ACES-Autopoietic Cognitive Edge-cloud Services-set out to rethink how Europe processes data where it matters most: at the edge. We’ve prototyped an edge-as-a-service (EaaS) that is self-aware, self-optimising, and sovereign by design. But the most important result may be what we learned together from you-our pilots, partners, and wider community-through the ACES surveys and interviews. Those insights now shape our next steps. This is your invitation to take part.

What you told us (and why it matters)

1) What matters most in an edge stack.

Across both supply and demand sides, five themes consistently rose to the top: (i) robust EaaS security (physical and logical), (ii) standardised service components (gateways, edge nodes, services, analytics/ML building blocks), (iii) edge analytics for real-time decisions, (iv) models that containerise/port assets from cloud to edge, and (v) the value of autopoietic functions (self-adjustment, orchestration, self-maintenance) for hands-off reliability. Suppliers placed especially high weight on autopoietic capabilities; demand-side respondents elevated security-including sector-specific security in verticals-and the practicalities of site management.

2) Shared vs. dedicated components.

A clear message emerged: suppliers are generally comfortable with shared (multi-tenant) compute, storage, and hardware-with dedicated options available as needed-while many users prefer more dedicated resources, especially for storage and internal communications. Both sides see some shared models for connectivity (edge-to-cloud and east-west between servers), but the demand for dedicated capacity is stronger on the user side, particularly where control, isolation, and performance guarantees are mission-critical.

3) What drives adoption.

Supply and demand converge on two primary drivers: closeness to the end user (better UX, lower latency) and OPEX optimisation through scalable architectures. Where views diverge: suppliers rank autopoietic features as a key driver for operational excellence, while users place comparatively higher emphasis on trust and transparent ownership of infrastructure and data.

4) What enables scale.

Both groups value high-performance connectivity and data-driven decision-making tools. Notably, blockchain, PID controllers, and probabilistic analysis are seen as strong enablers-especially by demand-side stakeholders who want autonomous, low-maintenance systems. Connectivity (e.g., 5G) matters, but many respondents emphasised that autonomy and AI-driven self-optimisation at the edge should carry more of the load.

5) What stands in the way.

Two cost items unite everyone: investment costs (hardware/software/HR) and migration costs are shared concerns. Then the emphasis splits: users stress skills and expertise gaps, orchestration/monitoring needs, and standardisation; suppliers highlight uncertainty around north–south data transfer costs and the need for better action tools for orchestrating distributed nodes. Security and privacy remain important, but many respondents treat them as table stakes; the operational and financial blockers are more acute.

6) Trade-offs to manage-not dodge.

There is broad agreement that edge and cloud are complementary: push real-time and east-west flows to the edge; keep resource-intensive analytics and long-term storage in the cloud. The open challenges are familiar yet solvable: CAPEX/OPEX at the edge, scalability for smaller deployments, and balancing local analytics with central workloads.

7) Performance priorities.

Both sides prioritise high availability, low latency, scalability/adaptability, and data security. Differences show up on autonomy (suppliers rate it higher) and UX transparency (users rate it lower than other operational factors)-another signal that practical performance and resilience are non-negotiable for adoption.

Bottom line: the market wants secure, standardised, low-latency edge platforms with predictable costs-and it wants them to self-manage as much as possible. That is exactly where ACES focuses: an autopoietic EaaS that reduces human toil, respects data sovereignty, and plugs cleanly into existing stacks.

What ACES will do next

We’re taking these findings into our final stretch by (1) deepening autopoietic orchestration and self-healing to shrink operating costs, (2) hardening vertical-grade security and governance paths demanded by pilots, (3) clarifying shared vs. dedicated deployment patterns and their SLAs, and (4) continuing to co-design with you through open activities and events.

Follow our updates on the ACES website News page for event announcements and demos.

Explore the CARO workshop (with our sister projects) to see how orchestration research is converging.

Three ways you can get involved-now

1) Join the ACES promotional webinar.

We’re hosting a public webinar to showcase what the autopoietic edge looks like in practice, including real pilot lessons and a candid discussion of trade-offs (costs, skills, standardisation). Register your interest and receive the invitation by subscribing via our “Join Us” page; we’ll send the date, agenda, and connection details directly to your inbox.

2) Bring your use case, shape the roadmap.

Are you evaluating dedicated vs. shared edge deployments? Struggling with orchestration across many sites? Want to keep east-west data local while staying cloud-compatible? Tell us where ACES can make the biggest difference in your environment-especially if you operate in sectors with strict compliance, low-latency needs, or complex multi-site operations. The survey evidenced exactly these pressure points (skills, cost predictability, orchestration), and we’re prioritising them in our final iteration.

3) Help us land the standards and the ecosystem.

Our interviews highlighted the importance of open interfaces and shared practices so that autopoietic features, telemetry, and policy controls can interoperate across vendors and verticals. If you contribute to standards bodies or run large heterogeneous fleets, we’d value your input on what “good” looks like for an edge that is both autonomous and accountable.

Talk to us

Project Coordination: Fernando Ramos (INESC-ID, IST, University of Lisbon)

Technical Coordination: Fred Buining (HIRO Microdatacentres)

Dissemination Coordination: Luca Alessandro Remotti (DataPower)

General info: info@aces-edge.eu

You can find the contacts-and follow our channels-on the ACES website’s Contacts page.

For general enquiries and press, writing to info@aces-edge.eu is the fastest way to reach the right person.

Why your participation matters

If the past decade was about cloud centralisation, the decade ahead will be about distributed intelligence: pushing cognition to where data is born, where milliseconds matter, and where sovereignty is non-negotiable. The ACES surveys underscore this pivot: users expect availability, latency, and security at the edge; they need predictable costs and tools that reduce manual ops; and they want clean interlocks with the cloud for heavy analytics and storage. That mix is precisely what ACES builds-an autopoietic edge that’s easier to run, cheaper to keep, and safer to trust.

So here’s our ask:

Subscribe today to get the promotional webinar invitation and future event notices.

Share your use case so we can validate ACES where it counts most.

Connect with the team via the Contacts page or info@aces-edge.eu to co-create pilots, contribute to standards, or discuss adoption paths.

Francesco Mureddu – DataPower Srl – Senior managing consultant