HORIZON-CL2-2026-01-TRANSFO-07 · The brief
Greenpulse — Participatory Understanding of Labour and Skills in Energy — would collect the meso-level evidence the call asks for: what workers, young people and communities in the energy sector actually know and do. It would then test the EU's competence frameworks against it.
Greenpulse is a proposal to Horizon Europe topic HORIZON-CL2-2026-01-TRANSFO-07 (“Fostering competences for the green transition”). Proposal submitted successfully · under evaluation. Everything here describes planned work, not a funded project.
The call
“Most available research on the green transition focuses on macro-level policies, overlooking how workers and communities can drive change.”
— Horizon Europe Work Programme 2026–2027, topic TRANSFO-07
The same text notes that “the availability of robust quantitative data is still limited”. Greenpulse would go where competences are made — firms, training sites, regional energy ecosystems, youth organisations and worker communities — and bring back the evidence the frameworks were drawn without.
The problem
The frameworks:
GreenComp, DigComp, EntreComp and LifeComp are built by expert consensus: GreenComp drew on about 75 experts. They give Europe a shared vocabulary, but they have never been tested against evidence from the places where competence is formed. Strictly, they are hypotheses about competence.
The training paradox:
Against a 60% adult-training target, only two adults in five train, and the most common reason is that they see no need. The research points to how jobs are structured and who is hired, not to the worker. The deficit sits in the productive system.
The green layer:
Europe labels green concepts in ESCO — 629 of them now, against 571 in 2022 — but the layer moves quietly between releases (110 added and 72 removed in one minor release), and no one checks it against real work. Europe's problem is not retreat but silent drift.
Greenpulse's move:
Greenpulse does not propose a rival framework. It tests the frameworks at the meso level, supplies the primary evidence they lack, and moves the question of “deficit” from the individual to the productive system, where the evidence points.
Energy, remade for security
Between 2023 and September 2026 Europe's energy system was remade by decisions taken for security rather than climate reasons. Cross-border energy systems are being deployed faster than the human capacity to operate, maintain and govern them.
AI at work
Generative AI reached 53% adoption in three years, faster than the PC or the internet. Required competence shifts from knowing how a machine is engineered to understanding what it is and how to work alongside it.
A legal duty without content
Since August 2026 national authorities enforce the AI Act's Article 4 duty on every deployer of AI to support its staff's AI literacy. No European instrument yet says what that means for an occupation, a sector or a workplace.
What already exists
What Greenpulse would add
SkillsPULSE measures how much mismatch there is and where. Greenpulse would explain why it reproduces where it does.
SkillsPULSE is not a partner; the link runs through joint briefs, shared workshops and cross-advisory contacts.
Overall objective
To take the pulse of the green transition where it is felt — in the energy sector, at the meso level — diagnose it, and recommend how to build the competences the twin green and digital transition needs. Greenpulse would make the competences that workers, young people and communities have and need traceable, teachable, translatable, transferable and updatable (T4U), each through a named deliverable.
Traceable
in the Observatory
D8.1, D8.10
Teachable
through the taught layer
D8.5
Translatable
through the xEET formats
D5.1
Transferable
through the meso–macro gap analysis
D8.2
Updatable
through versioned releases and ESCO change requests
D9.2
Five specific objectives
An empirically grounded, lifelong taxonomy of the competences workers and communities mobilise in the energy ecosystem, across five registers (OCTAB), tested against nature and climate adaptation as the intersectoral horizon.
→ Meso Competence Observatory (KER1)
Surveys, a narrative corpus (NECx) and a historical–archival strand (HACx) across six countries and two fieldwork rounds, triangulated with EU instruments and tagged to ESCO with the open-source ESCOX extractor.
→ Meso Competence Survey (KER3); Competence Equity Monitor (KER2)
Map how the two transitions' competences reinforce one another, and specify ontological literacy: the competences needed to work with increasingly agentic AI systems, for an occupation and a sector.
→ Digital–Green Competence Matrix, DG-CoM (KER5)
Programme guidance for education-and-training institutions (xGuidance) and a policy Roadmap (xRoadmap), a data pipeline back to GreenComp, the Union of Skills and the Pact for Skills via the Science-Meets-Politics conference, and policymaker–industry training protocols (xProPol).
→ Policy Roadmap and programme guidance (KER4)
Treat gender, disability and youth as perspectives that improve green-and-digital pathways, not only as groups that must benefit equally. Validated through the xEET formats, and extended to disadvantaged young people, including NEETs, with attention to the age at which orientation works best.
→ xEET replication modules; Equity Monitor metrics
Research-question coverage
The work programme lists ten research questions. Each is carried by one specific objective and answered by a named product.
Research design
What Europe prescribes, what training teaches and what workers and communities actually do, with the time depth of each. The distance between them — the prescribed/lived gap — is measured site by site and explained. Seven disciplines are held together by one shared taxonomy, one gap measure and joint calibration workshops before each fieldwork round; the project is led by the social sciences and humanities throughout.
Competence is made in situ.
Competence is treated neither as an individual attribute nor as a labour-market aggregate, but as something produced in real settings: firms, training sites, energy regions, youth and worker communities. Narrative is treated as primary data, and a narrative finding must pass a stated test; cases that resist the account are logged and reported.
What is prescribed and what is done never coincide.
From French work analysis (tâche prescrite / activité réelle, and the clinique de l'activité developed at CNAM): the gap between what frameworks prescribe, what training teaches and what workers actually do is where competence forms. Greenpulse measures that gap site by site and explains it with surveys, narratives and archives read together.
Competence as a way of being, not only a list.
Ontological, cognitive, technical, affective and behavioural. Greenpulse starts from the typology behind Europe's own conceptual work and adds two registers: the affective one, and the ontological one — the worker's relation to the object of work — stated as a hypothesis and tested where AI enters energy work.
Methods
Fieldwork runs in six countries (FR · BE · LV · PL · SE · ME) at 18 sites, three per country. Round 1 (months 8–14) opens two sites per country, 12 in all. After the Science-Meets-Politics conference (month 24), the Responsive Collection (ReCx, months 29–36) returns to the same skill ecosystems and site dyads with the findings and the policy response, and opens a third site per country. In the labels, an x after the name gathers from the field; an x before it returns evidence to the field or to policy.
The work plan is built on two pipelines, not on one cycle run twice. The evidence chain runs WP2 → WP3 → WP4 → WP7 → WP8, with WP6 computing the prescribed/lived distance. On that one chain, WP5 identifies and reduces the gap the field data carry, through original vocational formats measured pre/post on GreenComp; WP9 identifies and reduces the gap between those who decide and those concerned, opening it at the Science-Meets-Politics conference at month 24 and observing the response in Round 2. Both run on the same field protocol. WP1 and WP10 hold coordination and communication around them.
Sharing Circles and biographic interviews collect workers' and communities' own accounts at 18 sites: 54 Circles, about 350 participants and ~130–145 interviews. Based on the IRES Narrative Collection Method, developed and field-tested at Green Forum Sweden. The researcher's presence is varied across three conditions and its effect measured.
The Meso Competence Survey after Round 1, with ~1,800 respondents in six countries, and the Transition Choices Survey after the policy return, with ~700 in the last round – two instruments, not two waves. Cedefop's ETLS learning battery is reused verbatim, so results can be compared with European figures.
Dates each account against the site's own records and CNAM's archives, producing a site historiography and showing whether today's curriculum response times are fast, slow or without precedent.
A pre-registered test on the Digital–Green Matrix: does AI make competences converge across sectors, and only where the AI works in the workplace's own language? Reported whichever way it falls.
What the project would deliver
Each result has one owner: the partner whose science produced it, not necessarily the leader of the work package that hosts it.
Protocols
Integrate
Pathways to impact
TRANSFO-07 specifies three expected outcomes.
Expected outcome 1
Improved understanding of the competences needed for the green transition.
The Meso Competence Observatory, the survey datasets and the meso–macro gap analysis: for each site and territory, the distance between what the frameworks prescribe and what workers actually do, organised by skill-formation regime so findings travel by regime and not only by country. Narrative findings must pass a stated demarcation test, and cases that do not fit are logged and reported, not dropped. First Observatory release at month 32, final at month 46, with six country meso dossiers and a GreenComp × ESCO × O*NET concordance.
Expected outcome 2
Actionable advice for education-and-training programmes, in a lifelong perspective, with attention to gender and disability.
Two separate products: a policy Roadmap (xRoadmap) for policymakers and programme guidance (xGuidance) for training providers, the second co-produced with the providers at the study sites (month 44). Six country briefs on vocational systems (month 9), a re-runnable measure of how long training takes to respond to change at work, the Green Transition Academies moving from pilot to replication, and a Youth Orientation Module with a NEET pathway (month 28).
Expected outcome 3
Guidance on the green–digital interconnection.
The Digital–Green Competence Matrix crosses GreenComp and the ESCO green layer with DigComp 3.0, using both survey answers and coded narrative segments. On it, Greenpulse runs a pre-registered test: does AI make competences converge across sectors, and only where the AI works in the workplace's own language? The result feeds an AI-literacy specification for energy work (KER10) — content the AI Act's Article 4 duty now calls for, and which no European instrument yet specifies for an occupation or a sector.
Across all three · feedback into the EU skills infrastructure
Evidence goes back through the Science-Meets-Politics conference and a framework-feedback dossier to GreenComp, the Union of Skills and the Pact for Skills (D9.1), and through ESCO’s own contribution channels as documented change requests, with the GreenComp × ESCO × O*NET concordance submitted for listing among the portal’s crosswalks (D9.2). All of it is framed as submission and proposal: adoption remains the Commission’s decision.
≥12
open-access articles, plus 3 method papers, all preprinted
1 + 2
Science-Meets-Politics conference at month 24, then two working sessions (months 32 and 44)
6
policy briefs through the ENoP channel
6
country vocational-system briefs
3
Green Transition Academies of 30 participants each, each preceded by a youth orientation module
≥2
documented external replications of the Academy method
Advisory Board
If the project is funded, the General Assembly appoints a four-seat Advisory Board at month 3. It sits at months 8, 24 (in person, at the Science-Meets-Politics conference), 32 and 44, and its remit is deliberately narrow: to test the project’s claims against evidence produced outside it. Measurement and inclusion sit inside the consortium (UW; RSU’s occupational health and safety unit) and are not duplicated on the board. Members serve unremunerated and hold no beneficiary role.
Seat i
Narrative methods
Mary S. Morgan – London School of Economics; ERC Narrative Science programme (KER8)
Seat ii
EU skills policy
Lieve Van den Brande – formerly European Commission, DG Employment; Lifelong Learning Platform; TRANSVAL-EU (KER4)
Seat iii
Operational competence assurance in energy
Didier Ohayon – Inspectorate General for Nuclear Safety and Radiation Protection, EDF group; formerly IAEA (KER7)
Seat iv
VET systems and learning outcomes
Julian Stanley – formerly European Training Foundation; pre-accession VET and the Montenegrin case (RQ5, xGuidance)