Research cycles that once took years can now run in days. That was the message from a panel hosted by Huawei at the World Artificial Intelligence Conference (WAIC) 2026 in Shanghai, where researchers from the Beijing Academy of Artificial Intelligence (BAAI), the Chinese Academy of Sciences (CAS), the Shenzhen Loop Area Institute (SLAI) and Tsinghua University showed how AI is becoming an operational layer within science itself, rather than a bolt-on tool.
The Infrastructure Gap in R&D
Global spending on research and development continues to climb, yet the pace of scientific discovery has not moved at the same rate. Experimental cycles remain long, interdisciplinary collaboration faces structural barriers, and research workflows stay fragmented across instruments, disciplines and institutions.
The roundtable framed this as a systems problem as much as a scientific one — a question of engineering how to remove scientists from repetitive tasks and embed AI across the entire pipeline, from hypothesis to validation. That framing placed the emphasis on infrastructure design rather than isolated point solutions, a distinction all four institutions returned to throughout the session.
"Now, every time I see my postdoc using a pipette, repeating it hundreds or thousands of times, I think it's a waste. With humans, errors are added and efficiency is lowered. When you remove the human from it, everything accelerates and costs drop."— Yu Li, Director, State Key Laboratory of Membrane Biology, Tsinghua University
Automating Instrument Operation
SLAI, working with Suzhou National Laboratory, has built Owl·AuraID, a multi-agent system that automates experimental workflow from sample preparation through to data analysis. Rather than connecting to instruments through APIs, the system's agents operate equipment the way a person would — observing screens, clicking buttons and reading data directly, letting the platform link software agents, embodied scientific agents and instruments that were never built to communicate with each other.
The platform now covers six types of precision instruments. According to SLAI, it has cut the workload of crystal structure analysis by 50.6%, reduced morphological analysis time from nine minutes to 7.5 minutes, and lifted AI autonomous completion rates from 33% to 80%.
Some of the underlying hardware runs on chips developed by Huawei, with SLAI engineers working directly with Huawei's team to adapt a range of mainstream AI models to run on the platform.
"AI enables devices to interconnect, collaborate and optimise, freeing scientists from operations and redirecting their focus toward scientific insight."— Ouyang Wanli, Vice Dean, SLAI
Unifying Neuroscience Data
BAAI's contribution targets a different constraint: the lack of a shared data format across neuroscience recordings. Its Wujie·Brainμ1.0 model, described by BAAI as the world's first multimodal neuroscience foundation model, unifies electroencephalography, calcium imaging and neural probe signals within a single encoding framework, allowing previously incompatible recordings to be read together.
A study supported by the model appeared in Science in June 2026. According to BAAI, it showed for the first time that memory reactivation regulates sleep in both directions, with positive memories improving sleep quality and negative ones deepening fragmentation — findings with implications for treating sleep disorders linked to depression and anxiety.
The model was trained on more than 70,000 nights of sleep data and has run more than 12 months of automated analysis across partner laboratories. Its underlying data platform, Brain Token, pools open-source recordings from more than 20 sources alongside contributions from collaborating institutes and hospitals, currently spanning three species — humans, monkeys and mice.
"This trajectory is healthier and more promising than earlier LLM development, because demand is leading capability, and real-world application is driving model iteration."— Lei Bo, Researcher, BAAI