What remains after the project

Some results are operational, some are Beta, some are plans: a status-honest inventory of what is meant to continue after the project ends.

What is meant to continue after PoliRural+ ends in December 2026, and on what terms?

The open-source platform components (Hub, JackDaw, Magpie, Vulture, MapWhiteboard), which D4.2 describes as a functional, reusable Beta still to be consolidated; the free Advisor and Dashboard services; the public Atlas; the pilots' roadmaps to 2030–2040; and the open-call tools, 81 % of whose teams report continuing. D7.4 relies on a hybrid of public funding, private investment and subscription services, and fixes the final list of KERs in D7.8.

D4.2 calls the platform a functional and reusable Beta that still needs consolidation
D7.4: free and open source by default, a hybrid funding model, final KER list in D7.8
Pilot roadmaps run to 2030 and 2040; 81 % of open-call teams report continuing
The Atlas and this knowledge base keep the evidence public

Open in the interactive book · missions, mini-games and memory cards there

What remains after December 2026

main tellingdepth: standard3 min markdown

PoliRural+ ends on 31 December 2026. The table below lists what the deliverables expect to continue, with each item’s status as the sources word it.

What	Status in the sources	Terms for continuing
Platform: Hub/H4E, JackDaw, Magpie, Vulture, MapWhiteboard	Beta. D4.2: “a functional and reusable basis”, with integration, training and TRL progression still to be done	free, open-source cloud platform; Apache 2 recommended (D7.4)
Advisor and Dashboard of services	Advisor at TRL 7 in the Results Database (May 2025); D7.4: “the project is developing an Open Dashboard of services” for third-party services	free by default; D7.4 also recommends a subscription SaaS for governments and a freemium tier for business-intelligence tools
Atlas of best practices	public, 126 practices, edited by CCSS	open access
Open-call tools (e.g. LocalFarmsAI, Field2Fork AI Hub, SYLVAN-AI)	operational, with self-reported TRL 7–8 for Develop projects	81 % of surveyed beneficiaries report continuing (D6.3, self-reported)
Pilot roadmaps	RAP visions to 2030 and 2040, for the pilots to finalise in D7.8	regional and EU funds, partnerships

The terms behind the table

D7.4 builds sustainability on a hybrid model: public funding (EU and national grants, regional development funds), private investment, and subscription services for AI-powered advisory tools. Its conclusions name the risks plainly: slow bureaucracy, resistance to adoption, weak digital skills, funding gaps for early commercialisation and regulatory differences between regions. The final list of KERs, with their owners, is due in D7.8.

D4.2 asks the final release to focus on consolidation rather than new features: integration, simpler user pathways, a training and demonstration package, and continuous testing. Feedback from the pilots and open calls should become backlog items and test cases. Cooperation with FOCAL, BioClima, INTERESH, FAIR2Adapt and KijaniSpace shows that the same assets can serve other projects.

D6.3 adds three conditions for open-call outputs to last: open-source licensing, standard Model Context Protocol integrations, and continuous data feeds into shared European rural data spaces.

This knowledge base

This book belongs to the legacy too. It keeps the consortium’s knowledge readable for people and AI answer engines, and every claim in it points to a public deliverable. Chapter 4 seeds a local GeoAI knowledge base for Klatovsko that other regions can replicate.

The open question for 2027 is ownership more than technology: which partner keeps each service running, and in whose workflow.

Meet the results in person: Kilkenny and Prague, autumn 2026

another tellingdepth: intro1 min markdown

The results described in this chapter will be presented in public twice before the project closes.

Synergy Days 2026, Kilkenny (Ireland), 20–21 October 2026. The programme includes a PechaKucha session called “Agents in the Field”, with its own page at polirural-pechakucha.vercel.app. For background, the field validations of the JackDaw GeoAI agent are summarised under open-call results.

PoliRuralPlus final conference, CTU Prague, 20 November 2026. The closing conference of the project, held in hybrid format. CTU (CVUT) is the project coordinator.

For programme details, see the project website. These are the occasions to ask pilot teams and open-call beneficiaries the question this chapter leaves open: who keeps each result running after December 2026.

Check yourself: Which result would your region adopt first?

You coordinate development in a rural planning region that was not in PoliRural+. You have two staff and one year, and your council wants “something from that EU project” in use. Which do you pick first?

A) Copy a practice from the Atlas Find a practice in the Atlas of best practices from a region like yours and adapt it with local partners.

B) Roll out JackDaw to all planning staff now Give every planner the GeoAI chat and let them use it for daily work.

C) Run a small validation of JackDaw on your own data Load a handful of local datasets, test a fixed set of questions against GIS ground truth, then decide.

D) Procure the Advisor as a subscription service Budget for a commercial licence of the PoliRuralPlus Advisor and the Dashboard.

More than one answer can be defended. Choose one, then read why it works or where it breaks.

Feedback on each option

A) Defensible and cheap. The Atlas is public now, and every entry has a location, tags and results, so you can choose a practice tested in a context like yours, for example Vidzeme's indicator passports (prp-100) for municipal monitoring. Its limit is that the Atlas describes practices and does not run them; you still need local partners, and D6.3 counts trusted local intermediaries among the conditions for transfer.
B) Premature. D4.2 calls the platform a functional Beta whose final phase must still improve spatial query reliability and local context. The Enhance validations found answers to complex questions that were not reproducible (Bacău) and acceptance limited by data coverage and missing cited sources (AURA). Rolling it out to all staff before testing it on your own data invites disappointment.
C) The strongest first step if GeoAI is your goal. GeoEDUHealth Access – Bacău ingested eight local datasets, ran a 24-query battery in two languages against QGIS ground truth, and left a reusable 8-category testing protocol. You learn what works on your data before you commit, and your findings go back into the platform backlog as D4.2 asks.
D) Nothing to buy yet. D7.4 edition 1 describes subscription and SaaS models as potential routes, and says most KERs outside the business sphere will be free and open source. Which results are exploited, by whom and how is fixed only in D7.8 at the end of the project. A procurement plan built on D7.4's revenue tables is built on a plan, not on an offer.

Builds on: Key exploitable results (KERs)
Builds on: Atlas of best practices