Program

NEW [TECAN] Agentic AI in the Lab: From Reactive to Proactive lab management

Organizer: Tecan Trading AG

A Simultaneous interpretation provided

Agentic AI is transforming lab productivity - but what does that look like in practice? Join us for an engaging session with concrete use cases leveraging Agentic AI to move from reactive troubleshooting to proactive actions that can help prevent errors. See how autonomous AI agents can be used responsibly to automate key steps in life science research, uncover hidden patterns limiting lab throughput and enable scalability. The presented use cases will cover the scientific and operational impact, as well as the software and hardware foundations required to make them work. Whether you're in pharma, biotech, or clinical labs, you'll leave with concrete examples and insights to help you navigate your own AI journey.

Speakers: Hal Wehrenberg (Vice President, Head Global Services & Digital Innovation), Marco Ravot-Licheri (Head of Digital, Life Sciences Business)

[Acceleration Consortium]

Organizer: Acceleration Consortium

A Simultaneous interpretation provided

Part 1. Agentic Drug Discovery with Fármaco

The Matter Lab has pioneered the el agente ecosystem, a modular framework of AI-driven agents designed to autonomously orchestrate complex scientific workflows. Here, we introduce its most recent addition: Fármaco, an agent engineered specifically for computational drug discovery. By unifying predictive modeling with operational autonomy, Fármaco executes end-to-end in silico campaigns with minimal human intervention. Its core capabilities center on generative models for small molecule design paired with rapid binding affinity estimation. Equipped with a versatile toolkit to navigate large chemical spaces and formulate hypotheses, Fármaco is designed to effectively bridge the gap between early-stage computational design and downstream experimental validation, serving as a powerful new engine for discovery.

Part 2. Rapid Discovery of Potent CDK2 Inhibitors Using Boltz-2

Translating advanced in silico predictions into physical molecules remains a major bottleneck in drug development. To bridge this gap, we developed a streamlined platform coupling Boltz-2 affinity predictions with automated experimental validation, and showcased the platform for cyclindependent kinase 2 (CDK2), an important therapeutic target in oncology. We first curated a large virtual library of structurally diverse tetrahydroquinolines and utilized Boltz-2 to efficiently prioritize the most promising candidates. These in silico hits were then directly synthesized via a single-step aza-Diels–Alder approach. The resulting compounds underwent rapid evaluation through miniaturized chemiluminescence-based kinase assays. While automating the bioassay workflow at a 10 µL scale on Opentrons liquid handlers posed initial challenges, overcoming them ultimately drove significant improvements in precision and reproducibility. This seamless pipeline successfully yielded novel, single-digit micromolar inhibitors of CDK2.

Abstract (PDF)

Speaker: Sven Papidocha (University of Toronto)

[Material Handling] How Do You Handle Diverse Materials? Automating Chemistry & Materials Experiments: The Ground-Level Reality from Launch to Stable Operation

Organizers: Kazuki Ishizaki (The University of Tokyo), Shigeru Kobayashi (The University of Tokyo)

Photo/video recording: Not allowed · Social media posts: Not allowed

Language: Japanese

In the automation of chemistry and materials research, standardized instruments and protocols are not always readily available. Many challenges only become visible once you're actually in the field—device integration, sample handling, error recovery, and designing operations that keep experiments running without interruption. Drawing on case studies from areas such as polymer molding, thin films, alloys, and dielectrics, this session will explore the practical realities of building and operating automated experimental systems in material science. It will serve as a forum for discussing chemistry and materials automation from two key angles: where the real difficulties lie when bringing a system online, and what design choices and ingenuity are needed to sustain uninterrupted operation.

Speakers: Yuuki Asano (The University of Tokyo), Kensei Terashima (NIMS), Kazunori Nishio (The University of Tokyo), Shun Muroga (AIST)

This summary is machine-translated.

[Foundation Models] Scientific Foundation Models Powering AI for Science

Organizer: Haruka Ozaki (RIKEN)

Photo/video recording: Allowed · Social media posts: Allowed

Language: Japanese

Scientific foundation models are emerging as a powerful new research infrastructure underpinning AI for Science. This session invites researchers at the cutting edge of foundation model development to share real-world examples spanning model design, dataset construction, and application deployment. The discussion will cover the foundational technologies needed for AI to accelerate scientific research — and the road ahead.

Speakers: Seiji Akiyama (RIKEN), Koichi Higashi (ROIS), Ryosuke Kojima (RIKEN)

This summary is machine-translated.

[AI-Instrument Integration] Connecting AI Agents to Laboratory Instruments — Where Experimental Science Automation Stands Today and Where It's Headed

Organizer: Ryota Yamada (Science Aid, inc.)

Photo/video recording: Allowed · Social media posts: Allowed

A Simultaneous interpretation provided

Language: Japanese (English interpretation available)

While the use of AI agents is rapidly expanding in research and development settings, most applications remain confined to tasks that can be completed entirely on a PC — literature review, data analysis, and the like. This session turns the spotlight on the next frontier: connecting AI agents directly to laboratory instruments. Speakers who are actively working on this challenge will share live demonstrations of AI agent-driven instrument operation, and discuss everything from practical approaches to on-the-ground implementation to the broader outlook for experimental science automation.

Speakers: Hiroki Uchida (TECAN), Ryota Yamada (Science Aid, inc.), Teruaki Yamazaki (TECAN), Yiding Jia (Science Aid, inc.)

This summary is machine-translated.

[Automated Synthesis Operations] How Do You Actually Migrate All Those Organic Chemistry Experiments We've Been Slogging Through by Hand? The Long, Steep, and Painfully Difficult Road from Launch to Stable Operation

Organizer: Yuuya Nagata (NIMS)

Photo/video recording: Allowed · Social media posts: Not allowed

Language: Japanese

Automated synthesis systems are increasingly making their way into organic chemistry labs — but don't be fooled into thinking that simply installing a robot will instantly accelerate your research. Organic synthesis has evolved over decades, built on the accumulated experience and hands-on skill of researchers. As a result, a wealth of tacit knowledge — the kind that never makes it into written protocols — plays a decisive role in whether a synthesis succeeds or fails. On top of that, automated synthesis systems come with their own operational and process constraints. Adapting reagent supply and purification workflows to the equipment, and integrating data collection with experimental planning, adds further complexity. Simply swapping out traditional procedures for a machine is far from straightforward. This session focuses on the process of transitioning organic chemistry experiments to automated systems, and will explore the trial-and-error and real-world challenges encountered from the early stages of equipment introduction through to actual operation. We also believe these challenges are not unique to organic chemistry — they are shared across many experimental science disciplines. Researchers from all fields are warmly welcome.

Speakers: Kazushi Hayashi (Sagami Chemical Research Institute), Keisuke Nogi (Chugai Pharmaceutical Co., Ltd.), Hajime Shigemitsu (NIMS), Takuya Yamagata (Sagami Chemical Research Institute)

This summary is machine-translated.

[Life Science Lab Integration] Frontiers in Automated Workflow Design and Technology Integration for Life Science Experiments

Organizers: Takanori Uzawa (RIKEN), Atsushi Shibai (RIKEN)

Photo/video recording: Not allowed · Social media posts: Not allowed

Language: Japanese

Laboratory automation in the life sciences demands a wide variety of approaches depending on the target. This session presents the latest efforts in automating experimental operations across the life sciences, drawing on cases where automated systems have been built and deployed for diverse research targets — including microbial culture control, liquid dispensing, plant sampling, gut microbiome analysis, and protein sequence optimization. Each presentation will share the challenges encountered during implementation and the solutions devised along the way, with the session as a whole offering practical insights for advancing automation in real research environments.

Speakers: Miki Fujita (RIKEN), Kenta Katayama (Kobe University), Tetsushi Kawai (Chitose Laboratory Corporation), Kensuke Otsuka (CRIElPI), Hirotsugu Shiroma (OIST)

This summary is machine-translated.

NEW [Chemspeed] How Do You Connect Automation Instruments and AI? A Step-by-Step Path to a "Self-Thinking Experimental System" with Chemspeed

Organizer: AMR, Inc.

A Simultaneous interpretation provided

Language: Japanese (English interpretation available)

As AI technology becomes increasingly widespread, research labs are fielding a growing number of questions about one specific challenge: how to actually connect AI to existing automation instruments. Many researchers find themselves in a familiar bind — "We've set up the automation equipment, but the steps for integrating AI are unclear, and we don't know where to start." In this presentation, we use Chemspeed instruments — a highly scalable automation platform — as a concrete example to walk through practical approaches for lowering both the technical and psychological barriers to building an AI-integrated "self-thinking experimental system."

Speaker: Takato Toyama (AMR, Inc.)

This summary is machine-translated.

[Opentrons] Closing the gap between intent and execution in the lab

Organizer: Opentrons Labworks, Inc

A Simultaneous interpretation provided

AI is rapidly reshaping laboratory automation, moving beyond static scripting toward systems that can reason, perceive, and act in the physical world. In this session, Keith (Opentrons) explores what’s next for AI-enabled labs as large language models, computer vision, and robotics converge. He’ll examine how LLMs are evolving from tools that translate scientific intent into executable workflows into components of adaptive systems that plan experiments, interact with robotic platforms, and respond to changing requirements. He’ll also highlight the growing role of vision-based and multi-modal models for verification, real-time monitoring, and in-run adjustment. Looking ahead, Keith introduces emerging vision-language-action approaches that unify planning, perception, and execution, connecting digital intelligence with physical lab automation.

Speaker: Keith Tan (Opentrons Director of Channel Sales (APAC))

[AI Infrastructure] The Technical Foundations of AI for Science: "What Doesn't Change" in a Turbulent AI Era

Organizers: Tatsuro Ota (Chiba University), Taku Tsuzuki (Epistra Inc.)

Photo/video recording: Allowed · Social media posts: Not allowed

Language: Japanese

Behind the excitement surrounding AI for Science, how are computers actually working — and what limits are they up against? This session focuses on topics that are unlikely to go out of date even as the landscape shifts at a breakneck pace: the realities of data centers converting energy into information, the mechanisms driving continued hardware performance gains, and programming semantics for verifying code correctness. Bringing together experts in infrastructure, accelerators, and software, the session will offer a set of conceptual frameworks to light the way through the years ahead.

Speakers: Yuya Kawakami (SoftBank Corp.), Keigo Nishida (RIKEN), Takeshi Tsukada (Chiba University)

This summary is machine-translated.

[Cell Olympics] Defining the Events: Standard Tasks in Cell Manufacturing Processes

Organizers: Ayato Sugiyama (Astellas Pharma Inc.), Seiji Hori (VCCT Inc.)

Photo/video recording: Not allowed · Social media posts: Not allowed

Language: Japanese

Cell manufacturing is growing in importance across a wide range of fields — from regenerative medicine and cultivated meat/cellular agriculture to drug discovery. Yet the automation goals driving each of these fields are not necessarily the same. The metrics that matter most vary: quality, reproducibility, scale, cost, and throughput all carry different weight depending on the context. This session will examine the key distinctions between cell "culture" and cell "manufacturing," while exploring the challenges of automation in cell production. It will also look at reframing manufacturing steps — such as cell seeding, media exchange, and quality assessment — as discrete "automation tasks," with the aim of defining a shared set of challenges that researchers and industry can work toward together.

Speakers: Ryutaro Akiyoshi (Yokogawa Electric Corporation), Hiroko Hanzawa (Hitachi, Ltd.), Seiji Hori (VCCT Inc.), Atsushi Inoue (Cellafa Bioscience Inc.), Ikko Kawashima (IntegriCulture Inc.), Yoshiko Matsumoto (Eisai Co., Ltd.), Ayato Sugiyama (Astellas Pharma Inc.), Eigo Suyama (Chugai Pharmaceutical Co., Ltd.), Kazuhiro Takekita (Hyperion FoodTech Co.,Ltd.), Kenji Yoshimochi (NACALAI TESQUE, INC.)

This summary is machine-translated.

NEW [Ginkgo] How Autonomous Labs will replace the lab bench

Organizer: Ginkgo Bioworks

A Simultaneous interpretation provided

This seminar by Ginkgo Bioworks's CEO focuses on the real-world implementation of autonomous labs—drawing directly from active projects with OpenAI and Pacific Northwest National Laboratory (PNNL). Ginkgo's collaboration with OpenAI connected GPT-5 to Ginkgo's autonomous lab, improving cell-free protein synthesis reaction costs by 40% over the scientific state of the art. In parallel, the DOE commissioned over 100 integrated laboratory robots to enable national-lab scale automation for R&D as part of the White House's Genesis Mission to bring AI to science.

Speaker: Austin Che (Ginkgo Bioworks)

NEW [Unchained Labs] From Intent to Execution: Native AI-Driven Automation for Accelerated Scientific Discovery

Organizer: Unchained Labs

A Simultaneous interpretation provided

Laboratory automation has increased the speed and scale of experimentation, yet researchers are often still required to translate their ideas into complex workflows and scripts. What if scientists could simply describe their goals and let automation handle the rest? In this presentation, we introduce Stuntman, an AI-native automation platform that combines natural-language experiment design, automated execution, data analysis, and iterative optimization in a single environment. Attendees will see how AI can accelerate experimental progress across chemistry, biopharma, materials science, and energy applications. We will also show case studies of how existing Unchained Labs automation platforms have been utilized in both chemistry and biologics workflows.

Speaker: Andrew Stephenson, PhD (Sr Field Automation Scientist, Unchained Labs)

[Science Automation] Where We Stand on Fully Automating Scientific Research

Organizer: Haruka Ozaki (RIKEN)

Photo/video recording: Allowed · Social media posts: Allowed

Language: Japanese

Laboratory automation has advanced significantly, driven by progress in robotics, AI, and experimental data infrastructure. This session brings together researchers and developers at the forefront of automating the research process, who will share the latest developments in integrated systems spanning experimental planning, execution, and analysis. The discussion will explore where we currently stand on the path toward "full automation" of scientific research — and what lies ahead.

Speakers: Wataru Kumagai (RIKEN, NexaScience Inc.), Hafumi Nishi (Tohoku University, Ochanomizu University), Haruka Ozaki (RIKEN)

This summary is machine-translated.

[Robotics] Robotics Technologies Powering LA Today and a Vision for Tomorrow

Organizers: Makoto Jinno (Kokushikan University), Makoto Umeno (BAIKEIDO LLC, RIKEN)

Photo/video recording: Not allowed · Social media posts: Not allowed

Language: Japanese

This session provides an overview of robotics and mechatronics technologies in LA from the perspectives of hardware and system integration. It will showcase concrete examples of automation and efficiency gains through diverse robotic platforms, alongside an exploration of how the latest technologies will continue to evolve LA. Implementation essentials such as end-effectors and mechanism design will also be covered. Drawing on practical insights for researchers, developers, and users alike, the session will look toward the future of robotics in LA.

Speakers: Yasunori Nishihara (MAMEZO Co., Ltd.), Yasuyuki Oishi (Yamato Scientific co., ltd.), Yosuke Sawada (YAMAZAKI MAZAK CORPORATION.), Woosuck Shin (AIST)

This summary is machine-translated.