Wireless sensors and Yggio, the open operational layer for cities, property owners and transport operators. Your data, your operations, your sovereignty.

Distributed artificial intelligence across connected devices in the DAIS project
Home›Research Projects›DAIS - H2020-ECSEL-2020-2-RIA #101007273

DAIS - H2020-ECSEL-2020-2-RIA #101007273

Completed
Project lead: RISE Research Institutes of Sweden
← Back to Research Projects

Background

Artificial intelligence is increasingly moving from centralised cloud and data-centre environments towards connected devices at the edge. This makes it possible to process data closer to where it is generated, reducing latency and data transmission while enabling intelligent functions in increasingly small and resource-constrained devices.

DAIS (Distributed Artificial Intelligent Systems) was a large European research project focused on the technologies needed to make industrial-grade AI work across highly distributed edge environments. Rather than developing new AI algorithms, the project addressed the hardware, software, connectivity, security, and orchestration required to run existing AI algorithms across distributed systems.

The project brought together 48 partners from 11 countries to strengthen Europe’s capabilities in Edge AI, with privacy by design, energy efficiency, and self-organisation as fundamental design principles.

Bringing AI to the edge

Running AI directly on edge devices creates very different requirements from running it in the cloud. Devices may have limited processing power, memory, connectivity, and energy, while industrial applications can require real-time operation, reliability, security, and protection of sensitive data.

DAIS developed both hardware and software for Edge AI, ranging from specialised electronic components and low-power sensor technologies to software capable of running efficiently on resource-constrained devices.

A key objective was to significantly reduce the energy consumption, physical size, and amount of data exchanged between edge devices and the cloud while still enabling intelligent processing and learning close to the source of the data.

Distributed AI from edge to cloud

Many AI applications cannot operate entirely at one location. Processing may be distributed between sensors, embedded devices, gateways, edge computers, fog infrastructure, and cloud platforms depending on available resources and application requirements.

DAIS therefore investigated how distributed AI components could be securely integrated and coordinated across edge, fog, and cloud environments. This included orchestration of AI tasks, distributed security and privacy mechanisms, networking, and communication between different parts of the system.

The result was a distributed AI framework designed to allow components, platforms, and services from different providers to exchange data and AI analysis results while remaining part of a larger interoperable system.

From components to real-world demonstrators

DAIS combined technology development with large-scale integration and validation. Hardware, software, AI services, sensors, and platforms developed across the project were brought together to demonstrate that distributed AI could operate across organisational and technological boundaries.

The technologies were demonstrated in three physical domains: an industrial location, a building, and a city. These environments were used to test both vertical integration within individual applications and horizontal integration where technologies and AI results could be shared between different systems.

By validating the same underlying technologies across multiple domains, DAIS demonstrated how distributed AI architectures could support applications in digital industry, digital life, transport, mobility, and other connected environments.

Sensative’s role

Sensative had a central integration role in DAIS, leading the project’s work on final demonstrators and the sub-cluster responsible for coordination and orchestration of distributed AI.

Yggio was used as a horizontal middleware and data-broker platform to connect devices, AI components, platforms, and services developed by different project partners. This allowed data and AI analysis results to be shared between otherwise independent systems and provided a practical environment for validating the DAIS distributed AI architecture.

The project established important foundations for Sensative’s work with distributed Yggio architectures, where intelligence, applications, and data processing can be distributed between devices, edge infrastructure, and cloud environments.

Distributed AIEdge AIArtificial IntelligenceEdge ComputingIndustrial IoT
Other participants (48)
  • ALMENDE BV
  • DANMARKS TEKNISKE UNIVERSITET
  • SIRRIS HET COLLECTIEF CENTRUM VAN DE TECHNOLOGISCHE INDUSTRIE
  • TECNOLOGIAS SERVICIOS TELEMATICOS Y SISTEMAS SA
  • INSTITUT JOZEF STEFAN
  • INSTITUT MIKROELEKTRONICKYCH APLIKACI S.R.O.
  • JOTNE EPM TECHNOLOGY AS
  • PDM E FC PROJECTO DESENVOLVIMENTO MANUTENCAO FORMACAO E CONSULTADORIA LDA
  • EXPERT ANALYTICS AS
  • Step Solutions AS
  • COSYLAB LABORATORIJ ZA KONTROLNE SISTEME DD
  • TPV trzenje in proizvodnja opreme vozil d.d.
  • INSTITUTO DE TELECOMUNICACOES
  • NXP SEMICONDUCTORS NETHERLANDS BV
  • ANYWI TECHNOLOGY BV
  • DANFOSS POWER ELECTRONICS AS
  • TECHNICKA UNIVERZITA V LIBERCI
  • VYSOKE UCENI TECHNICKE V BRNE
  • I. LECO
  • CENTRICA BUSINESS SOLUTIONS BELGIUM
  • Octavic PTS IVS
  • NXP SEMICONDUCTORS GERMANY GMBH
  • FRAUNHOFER GESELLSCHAFT ZUR FOERDERUNG DER ANGEWANDTEN FORSCHUNG E.V.
  • NUROMEDIA GMBH
  • GIESECKE+DEVRIENT MOBILE SECURITY GMBH
  • ESC AEROSPACE GMBH
  • UNIVERSITAET ZU LUEBECK
  • Eventigrate
  • VESTEL ELEKTRONIK SANAYI VE TICARET ANONIM SIRKETI
  • SESTEK Ses ve Iletisim Bilgisayar Teknolojileri Sanayi ve Ticaret A.S.
  • Akim Metal
  • SCHNEIDER ELECTRIC ESPANA SA
  • FUNDACION CENTRO TECNOLOXICO DE TELECOMUNICACIONS DE GALICIA
  • UNIVERSIDAD DE CANTABRIA
  • UNIVERSIDAD DE GRANADA
  • TECHNISCHE UNIVERSITEIT EINDHOVEN
  • SEVEN SOLUTIONS SL
  • INSTITUTO TECNOLOGICO DE INFORMATICA
  • TECHNISCHE UNIVERSITEIT DELFT
  • STICHTING IMEC NEDERLAND
  • INNATERA NANOSYSTEMS BV
  • SENSATIVE AB
  • TIETO SWEDEN AB
  • GREENSPHERE UNIPESSOAL LDA
  • Koala Tech Unipessoal LDA
  • Arctos Labs Scandinavia AB
  • BEYOND VISION - SISTEMAS MOVEIS AUTONOMOS DE REALIDADE AUMENTADA LDA
  • MAELARDALENS HOEGSKOLA
Funded by
EU grant agreement 101007273 (CORDIS) (opens in a new tab)

Research enquiries

cto@sensative.com