INSTITUTE OF AUTOMATION CHINESE ACADEMY OF SCIENCES

Their core work

CASIA is one of China's flagship research institutes for artificial intelligence, pattern recognition, robotics and automation, operating under the Chinese Academy of Sciences in Beijing. Their work spans computer vision, biometric identification, brain-inspired computing, intelligent robotics and neural system modelling, with strong capability to move algorithms from lab prototypes to VLSI hardware and real-world systems. In the H2020 context they served as a non-EU partner in MSCA-RISE exchange networks, hosting European researchers and contributing Chinese expertise in vision, forensics, and bio-inspired perception. For European partners they offer access to large-scale AI research capacity, complementary datasets, and a bridge into the Chinese robotics and computer vision ecosystem.

What they specialise in

How they've shifted over time

Their earliest H2020 activity (2016-2017) focused on applied computer vision — forensic identification in IDENTITY and robot perception in SMOOTH — a classical automation and pattern-recognition profile. From 2018 onward with ULTRACEPT the work shifted toward biologically inspired perception, combining insect visual pathways, thermal imaging and VLSI implementation for automotive safety. The direction of travel is a clear move from conventional vision pipelines toward neuromorphic, multi-modal perception aimed at safety-critical systems.

Moving toward neuromorphic, multi-modal perception for safety-critical automotive and robotic applications, making them a strong future partner for autonomous systems and edge-AI hardware consortia.

How they like to work

CASIA enters European consortia exclusively as a third-party partner in MSCA-RISE staff exchange networks, never as coordinator or formal beneficiary. They engage in mid-sized international consortia, having worked with 41 different partners across 19 countries — a high partner-per-project ratio that points to a hub role bringing researchers in and out rather than locking in with the same few groups. European partners should expect strong research contribution and researcher mobility, but contractual leadership will sit with an EU coordinator.

They have collaborated with 41 unique partners across 19 countries over just three projects, a high ratio that reflects the broad international membership typical of MSCA-RISE networks. The reach is clearly global, with a strong European anchor combined with China as the hosting base.

What sets them apart

CASIA is the single largest automation and AI research institute in China, which makes them almost unique among H2020 third-party partners as a gateway into Chinese AI research, datasets and robotics labs. They combine algorithmic depth with hardware design (VLSI) and large-scale application testing — a mix few European university groups can match. For a European coordinator, partnering with CASIA means researcher exchange, complementary non-EU datasets and the scale of a national academy institute.

Highlights from their portfolio

• ULTRACEPT
  Their longest and most technically distinctive project (2018-2024), fusing insect-inspired neural models, thermal imaging and VLSI hardware for vehicle collision avoidance.
• IDENTITY
  Brought CASIA into a European multimedia forensics network, applying their computer vision expertise to biometric identification.
• SMOOTH
  Extended their robotics capability into a high-stakes application domain — smart firefighting robots — demonstrating cross-over from automation to emergency response.