AI knowledge flow · research reuse

Where does knowledge from funded AI research travel?

The index connects German participation in EU AI projects to reported research outputs and uses exact DOI and OpenAlex identifiers to observe where those works are cited next.

Exact identifiers, not name matchingFull baseline and OpenAlex run kept separateNo generative AI model

Exact OpenAlex probeCollector checked 16 Jul 2026, 22:16Source data through 11 May 2026Tracking since 16 Jul 2026Revision 4

Full cohort · CORDIS + OpenAIRE

The reliable baseline

Full baseline checked

Project and output aggregates come from the fully checked transfer cohort. They are not extrapolated from the OpenAlex run.

reported research outputs in the full cohort: 50,609
outputs with at least one DOI: 35,926
unique DOI identifiers in the full cohort: 61,591
reported OpenAIRE citations: 716,634

Exact OpenAlex probe

The observed knowledge flow

Bounded exact probe

The first release resolves every DOI in the public output sample exactly against OpenAlex. The sample is bounded and not representative.

DOI outputs in the OpenAlex run: 477
DOI identifiers in the OpenAlex run: 784
exactly resolved in OpenAlex: 461
observed output-to-citation edges: 516

Selection rule: every DOI-bearing record in the public transfer sample; no extrapolation

Search projects, outputs or institutionsFramework programmeCountry of citing institutionInstitution type

180 results

Output → citing work → institution

Exactly observed knowledge paths

Choose an output to see which works reference it and which institutions OpenAlex reports for them. The graph shows at most five outputs and eight recent citing works per output; the metrics count every observed edge in the active OpenAlex run. Person and author data are not published.

Exact OpenAlex probeBounded exact probe

Funded output

Citing institutions

Fraunhofer Institute for Telecommunications, Heinrich Hertz InstituteDE · 3 Citing works

Nanyang Technological UniversitySG · 3 Citing works

Technische Universität BerlinDE · 3 Citing works

Berlin Institute for the Foundations of Learning and Data2 Citing works

Chongqing University of Posts and TelecommunicationsCN · 2 Citing works

Freie Universität BerlinDE · 2 Citing works

Lanzhou UniversityCN · 2 Citing works

Latest publicly displayed relations · at most eight works per output
Citing works	Citing institutions	Country
Enhanced RGB-D feature extraction for 6D pose estimation	China Datang Corporation (China)	CN
6D pose estimation method based on hybrid attention mechanism and vector-based local consistency enhancement	Chongqing University of Posts and Telecommunications	CN
RayPose: Hand Joint Ray Aggregation for 6DoF Object Pose Estimation	Dalian University of Technology	CN
Occlusion-resilient pose estimation of textureless components in cluttered environment and its implementation in robotic bin-picking	Indian Institute of Technology Kharagpur	IN
Pose measurement for shipborne aircraft autonomous landing via onboard visual-inertial-altitudinal data fusion	Institution not reported	unknown
Enhanced RGB-D feature extraction for 6D pose estimation	Ningde Normal University	CN
PoseIDON: 6DoF pose estimation with foundation model features for marine sediment burial mapping	Scripps Institution of Oceanography	US
Enhanced RGB-D feature extraction for 6D pose estimation	Tsinghua University	CN
Object pose estimation for upper-limb prostheses grasping.	Université de Bordeaux	FR
PoseIDON: 6DoF pose estimation with foundation model features for marine sediment burial mapping	University of California San Diego	US
SynBag: Synthetic Training Data for Autonomous Grasping of Regolith Bags in the Lunar Environment	Western University	CA
Investigating the internal structure of X ( 6900 ) in the 2 J / ψ decay channel	Guangxi University	CN
Discovering an unquenched dynamics mechanism for charmonium scattering	Hunan Normal University	CN
Discovering an unquenched dynamics mechanism for charmonium scattering	Institute of Modern Physics	CN
Two-charmonium scattering with the quark Pauli-blocking effects	Japan Proton Accelerator Research Complex	JP
Discovering an unquenched dynamics mechanism for charmonium scattering	Lanzhou University	CN
Correction to the chromoelectric interaction energy of the fully heavy tetraquark state	Lanzhou University	CN
Correction to the chromoelectric interaction energy of the fully heavy tetraquark state	Lanzhou University of Technology	CN
Two-charmonium scattering with the quark Pauli-blocking effects	Nagoya University	JP
Discovering an unquenched dynamics mechanism for charmonium scattering	Nanjing Normal University	CN
Correction to the chromoelectric interaction energy of the fully heavy tetraquark state	Northwest Normal University	CN
Two-charmonium scattering with the quark Pauli-blocking effects	Obayashi (Japan)	JP
Two-charmonium scattering with the quark Pauli-blocking effects	RIKEN Nishina Center	JP
Two-charmonium scattering with the quark Pauli-blocking effects	Showa Pharmaceutical University	JP
Systematic study of exotic 1 − + tetraquark spectroscopy	Suranaree University of Technology	TH
Systematic study of exotic 1 − + tetraquark spectroscopy	Suranaree University of Technology	TH
Two-charmonium scattering with the quark Pauli-blocking effects	The University of Osaka	JP
All-charm tetraquarks at hadron colliders: A high-precision fragmentation perspective	Universidad de Alcalá	ES
Multimodal fragmentation of all-heavy pentaquarks: Uncertainty-aware predictions for hadron colliders	Universidad de Alcalá	ES
Correction to the chromoelectric interaction energy of the fully heavy tetraquark state	Yili Normal University	CN
Capacity analysis of OMA-PAS and NOMA-PAS	Beijing Institute of Technology	CN
Capacity analysis of OMA-PAS and NOMA-PAS	Chongqing University	CN
Capacity analysis of OMA-PAS and NOMA-PAS	Chongqing University of Posts and Telecommunications	CN
Capacity analysis of OMA-PAS and NOMA-PAS	King Abdullah University of Science and Technology	SA
Channel Estimation for Pinching Antennas Systems using Deep Learning	King Fahd University of Petroleum and Minerals	SA
Capacity Characterization of Pinching-Antenna Systems	Kyung Hee University	KR
Hybrid Pinching Antenna Systems: Architecture and Beamforming Design	Memorial University of Newfoundland	CA
Resource allocation for multi-user pinching-antenna system	Nantong University	CN
Uplink and Downlink Communications in Segmented Waveguide-Enabled Pinching-Antenna Systems (SWANs)	Nanyang Technological University	SG
Capacity Characterization of Pinching-Antenna Systems	Nanyang Technological University	SG
Hybrid Pinching Antenna Systems: Architecture and Beamforming Design	Nanyang Technological University	SG
Self-Supervised graph attention–based antenna activation for pinching antenna systems under uncertainty	National Institute of Technology Tiruchirappalli	IN
Effective Spectral Efficiency Maximization for Directional Pinching-Antenna-Assisted Multi-User MIMO Systems	Northeastern University	CN
Uplink and Downlink Communications in Segmented Waveguide-Enabled Pinching-Antenna Systems (SWANs)	Queen Mary University of London	GB
Channel Estimation for Pinching Antennas Systems using Deep Learning	South East Technological University	IE
Hybrid Pinching Antenna Systems: Architecture and Beamforming Design	Southeast University	CN
Self-Supervised graph attention–based antenna activation for pinching antenna systems under uncertainty	Sri Ramakrishna Institute of Paramedical Sciences	IN
Uplink and Downlink Communications in Segmented Waveguide-Enabled Pinching-Antenna Systems (SWANs)	University College Dublin	IE
Capacity Characterization of Pinching-Antenna Systems	University College Dublin	IE
Uplink and Downlink Communications in Segmented Waveguide-Enabled Pinching-Antenna Systems (SWANs)	University of Hong Kong	HK
Capacity Characterization of Pinching-Antenna Systems	University of Hong Kong	HK
Channel Estimation for Pinching Antennas Systems using Deep Learning	Waterford Institute of Technology	IE
Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces	BASF (Germany)	DE
Explainable AI for time series via Virtual Inspection Layers	Berlin Institute for the Foundations of Learning and Data	unknown
From Hope to Safety: Unlearning Biases of Deep Models via Gradient Penalization in Latent Space	Berlin Institute for the Foundations of Learning and Data	unknown
Explaining deep learning for ECG analysis: Building blocks for auditing and knowledge discovery	Carl von Ossietzky Universität Oldenburg	DE
Explaining deep learning for ECG analysis: Building blocks for auditing and knowledge discovery	Charité - Universitätsmedizin Berlin	DE
From classification to segmentation with explainable AI: A study on crack detection and growth monitoring	École Polytechnique Fédérale de Lausanne	CH
Explainable AI for time series via Virtual Inspection Layers	Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute	DE
From Hope to Safety: Unlearning Biases of Deep Models via Gradient Penalization in Latent Space	Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute	DE
Explaining deep learning for ECG analysis: Building blocks for auditing and knowledge discovery	Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute	DE
Explainable AI for time series via Virtual Inspection Layers	Freie Universität Berlin	DE
Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces	Freie Universität Berlin	DE
Exploring Dataset Bias and Scaling Techniques in Multi-Source Gait Biomechanics: An Explainable Machine Learning Approach	Friedrich-Alexander-Universität Erlangen-Nürnberg	DE
Exploring Dataset Bias and Scaling Techniques in Multi-Source Gait Biomechanics: An Explainable Machine Learning Approach	Helmholtz Zentrum München	DE
Explainable Deep Neural Networks for Predicting Mutated Patterns in SARS-CoV-2 Variants with Geographic Analysis	Indian Institute of Information Technology Design and Manufacturing Jabalpur	IN
Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces	Korea University	KR
Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces	Max Planck Institute for Human Cognitive and Brain Sciences	DE
Explaining deep learning for ECG analysis: Building blocks for auditing and knowledge discovery	Physikalisch-Technische Bundesanstalt	DE
Challenges in explaining deep learning models for data with biological variation	Technical University of Denmark	DK
Explainable AI for time series via Virtual Inspection Layers	Technische Universität Berlin	DE
From Hope to Safety: Unlearning Biases of Deep Models via Gradient Penalization in Latent Space	Technische Universität Berlin	DE
Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces	Technische Universität Berlin	DE
A combinatorial ABC algorithm for AoI minimization with reliable data collection in UAV-assisted clustered IoT networks	Erciyes University	TR
AI-Driven Digital Transformation and Sustainable Logistics: Innovations in Global Supply Chain Management	Ferdowsi University of Mashhad	IR
Neural networks for socio-labor regulation: a neuromorphic approach to human-centric AI in urban economies	Heilongjiang University	CN
An Adaptive Governance-Centric MLOps Framework for Risk-Tiered Control and Continuous Assurance of Responsible AI in High-Stakes Domains	Institution not reported	unknown
Towards embodied AI in manufacturing: Review, Evaluation, and Future directions	Institution not reported	unknown
Autonomous vehicles and Quality 5.0: a conceptual paper on the role of artificial intelligence providers	Parthenope University of Naples	IT
Neural networks for socio-labor regulation: a neuromorphic approach to human-centric AI in urban economies	Peter the Great St. Petersburg Polytechnic University	RU
Research Landscape of Industry 5.0: A Bibliometric Analysis and Thematic Synthesis	Silesian University of Technology	PL
Innovación, IA y diseño centrado en el ser humano: el rol del diseño industrial en la quinta revolución industrial	Universidad Técnica de Ambato	EC

Choose output

Performance Analysis of Pinching-Antenna Systems

6G-XCEL

observed output-to-citation edges: 37
OpenAIRE citation signal: 18
highest OpenAlex citation count of a resolved version: 39
first observed citation: 01 Jan 2025

Open source ↗

Method & identity rules

Knowledge flow without invented bridges

The collector connects sources only through compatible official or persistent identifiers. Missing assignments remain visibly unknown.

1
Define the project cohort exactly
CORDIS project codes and reported German participation define the AI cohort; they do not prove German authorship of every output.
2
Take outputs from OpenAIRE
Research products are connected only through the project relation reported by OpenAIRE. Full-cohort aggregates remain separate from the OpenAlex run.
3
Preserve every DOI relation
Every valid DOI of an output is processed separately. The collector picks no supposed primary DOI and matches no titles or person names.
4
Resolve references backwards
A citing work counts only when its OpenAlex reference list contains a resolved target work. Institutions come from reported affiliations.

Collection, linking, classification and presentation use no generative AI model.

How to read the figures correctly

Citations evidence scholarly attention or reuse, not quality or economic or societal impact.
CORDIS proves German participation in the project. It does not prove that a German project participant authored a linked output.
The OpenAlex flow analysis is a bounded, non-representative probe of the public transfer sample. Full-cohort baselines are shown separately.
Countries, sectors and institutions are full-counted from reported OpenAlex affiliations and may overlap. Missing affiliations remain unknown.
Sources add, correct and deduplicate records retrospectively. Recent months are therefore marked provisional.

Sources, licences & individual coverage status

Each source shows its check status. The upstream cutoff is stated separately from the collector run time; bounded samples are never extrapolated.

European Commission · CORDISavailable

953

Licence / reuse: European Commission reuse policy

checked 16 Jul 2026, 22:16Open source ↗

OpenAIRE Graphavailable

50,609

Licence / reuse: CC BY 4.0

checked 16 Jul 2026, 22:16Open source ↗

OpenAlexbounded sample

708

Licence / reuse: CC0

checked 16 Jul 2026, 22:16Open source ↗

ROR identifiers via OpenAlexavailable

1,010

Licence / reuse: CC0

checked 16 Jul 2026, 22:16Open source ↗

Frequently asked questions

What does AI knowledge flow measure?

It observes which works cite outputs from EU-funded AI projects and which countries, sectors and institutions occur in the reported affiliations of those citing works. This is a signal of scholarly attention and reuse.

What coverage does the OpenAlex analysis have?

The observed-flow card states the active scope. A bounded probe is never extrapolated. A complete crawl processes the configured DOI cohort; unresolved DOIs remain reported as gaps.

Does the index prove German authorship or impact?

No. CORDIS proves German project participation, not authorship of every output. A citation also proves neither quality nor economic or societal impact.

Why can one output have several DOI and OpenAlex links?

OpenAIRE can report multiple persistent identifiers for one output, such as a preprint and journal version. The collector preserves this many-to-many relation instead of arbitrarily choosing one primary identifier.

Does the index publish author names?

No. Work, institution and ROR identifiers are sufficient for the knowledge-flow analysis. Person, author and ORCID data are excluded from the public dataset.

Does the collector use generative AI?

No. It uses versioned rules, exact identifiers, API responses and schema validation. There is no language model in the collector and no generated assignment.

Want to connect research, funding or market data reliably?

We build traceable data products with automated collectors, stable identity rules and verifiable provenance.

Discuss a project View all tools

Related tools

Useful next steps

These tools complement the current result.

View all tools

IndexAI research & transferWhich research products does OpenAIRE link to publicly funded AI projects? Publications, datasets, software, access metadata and citation signals, joined to CORDIS projects by exact grant identifiers.IndexAI innovation pathFollow evidence-backed links from EU funding to reported research outputs—and see exactly where patents or procurement cannot yet be joined safely.IndexGerman AI influence & public moneyWho discloses AI lobbying, which matters move through parliament, and where is EU funding allocated? A daily-checked source graph built from Lobbyregister, DIP and CORDIS.