Takanori Yamamoto
山本尚範Project Lead
Head, Department of Emergency and Medical Intensive Care, Nagoya University Hospital. Supervising DMAT physician. Aichi Prefecture Disaster Medical Coordinator.
AIMER · AI FOR MEDICAL EMERGENCY RESPONSE
Care mustnot stop.
Information in time to decide.
In the chaos of a disaster, decisions are being made to save one more life.
We connect scattered information and surface the risks not yet seen.
AIMER (AI for Medical Emergency Response) supports life-saving decisions with advanced AI and data.
AI-generated concept visual.
The figures on screen are fictional.
Google.org
Google.org Impact Challenge: AI for Government Innovation 2026Selected project
Official announcement (opens in a new tab)MISSION Why we do this01
Caused by information that arrives too late:preventable disaster deaths, and disaster-related deaths.0is the target.We design the system to reach it.
In a disaster medical headquarters, information about hospital status, patient transfers, staffing, fuel and medical gas arrives without pause, and decisions become hard to make. We deliver that information in a form that arrives in time for the decision, so that experts can decide with evidence in hand.
516/ 744 lives
Lives lost after the shaking stopped,
Noto Peninsula earthquake (2024)*1
9channels
Parallel information channels in one disaster response.
Reconciled by hand*2
On the ground, people save lives by improvising.
But improvisation alone does not spread.
AI only gathers and presents the data.
Experts make the decision on what it shows.
About the figures and the quote
*1 "516 / 744" is carried over from the supplied copy (LP v0.2). The reference date, area, scope and original source need to be confirmed. The figure does not represent deaths caused by delayed information, nor the number of lives this AI could save.
*2 "9 channels" follows the count used in the supplied materials: EMIS, D24H, J-SPEED, Join, Google Forms, Google Drive, telephone, fax, and paper or Excel. The actual number of channels varies by site and by how finely they are classified.
*3 The wording and attribution of the quote follow the revised copy supplied to us. Before publishing it as a quotation, confirm the record of the remark and permission to publish.
OBJECTIVES What we aim for02
Information
in time to decide,
built into a system
rooted in the field.
This project does not end with AI research. Alongside the technology, we work on four more fronts: society, design, people and validation. Together they aim at a system that is actually used, trusted and useful in the field of disaster medicine, and that is carried forward to the next disaster.
2026.10 – 12Technical and field validation. Replay headquarters records from past disasters and test inside exercises.
2027 –Define requirements from the validation results and develop together with the field.
3 YEARSPut it into practice as the system that supports headquarters decisions in the next disaster.
- TechnologyTECHNOLOGY
Gather the scattered records of an emergency and turn them into proposals.
In a disaster, all kinds of unsorted information pours in. Demand for fuel, water and medical gas, and patient transfers: this information is tied to lives. Building on records of past disasters and simulations of future ones, we provide information that is fast and reliable.
- DesignDESIGN
Deliver every proposal to the decision-maker with its evidence and alternatives.
Each AI proposal arrives with its evidence, its confidence and the alternatives. Approving, editing or returning it is a human decision. We design for people to decide, not for AI to decide everything. The system coexists with existing systems and adds no new work for the field.
- PeoplePEOPLE
Trust in the field, and seeds for the future.
Experts in disaster medicine and AI join forces as one team, work with partners and the practitioner community, and earn the trust of the field. We also invest in developing the next generation, building a sustainable organization alongside the system itself.
- SocietySOCIETY
Draw on past and present knowledge to take on a new system.
We organize the records accumulated in past disasters and the knowledge used in today's exercises into forms that research can use. Wherever possible, we build working relationships with government and medical institutions, so that the system is validated in real exercises and operations, not only on paper.
- ValidationVALIDATION
Put usefulness and safety first.
On the road to real-world use, we give the highest priority to validating effectiveness and evaluating safety. Working with disaster medicine practitioners, we assess where the latest AI technology can contribute, and whether it is effective and safe.
These five objectives are a public-facing paraphrase of the Objectives in the OKRs (Objectives and Key Results) set for the Google.org program. The numerical targets will be tested during validation and compiled in a validation report. Participation in exercises and use of records proceed with the approval of the organizations concerned and ethics review.
TEAM Who we are03
Experts in disaster medicine
and in geospatial data and AI, as one team.
An emergency physician who has led medical teams at disaster headquarters, a geospatial researcher who reads disasters through mobile data, a researcher in medical-data governance and AI design, a researcher advancing AI for the public good through human behavior modeling, and a public-sector partner responsible for disaster medicine in Aichi Prefecture.
Mariko Shibasaki
柴崎真理子Deputy Project Lead / AI for the public good and human behavior modeling
Visiting Researcher, Center for Spatial Information Science, The University of Tokyo.
Masanobu Tsuda
津田雅庸Public-Sector Implementation Partner
Aichi Prefecture Disaster Medical Coordinator. Aichi Medical University. Aichi DMAT.
Atsushi Koshio
小塩篤史System Development / AI Agents
Professor, Faculty of Engineering, Reitaku University. CTO, HYPER CUBE Inc.
Ryosuke Shibasaki
柴崎亮介Geospatial Analysis & Geospatial AI
Vice President and Professor, Reitaku University. Professor Emeritus, The University of Tokyo.
GOOGLE.ORG Support team
Joel Cheong
Google AI Coach
Yuki Ihara
Google Project Success Manager
Isao Tezuka
Google Executive Sponsor
Nagoya UniversityReitaku UniversityHYPER CUBE Inc.
Names, titles, roles and photos are based on supplied copy. Confirm the current wording and permission to publish before release.
BUILD WITH US Collaboration & media inquiries04
Before the next disaster,
let's build it together.
FROM NAGOYA & AICHI.
TO COMMUNITIES ACROSS JAPAN.
This project starts in Nagoya and Aichi, but large-scale disasters are a challenge for all of Japan.
Working with people in many regions, we want to bring AI that supports disaster medicine, and training for it, into real-world use.
That takes knowledge from the front line of disaster medicine, data and technology of many kinds, and validation in the field.
If the project interests you, please get in touch.
CONTACTCollaboration
Collaboration
& media inquiries
Wherever you are and whatever your role,
let's start with a conversation.
- 01 Enter
- 02 Review
- 03 Send
This form is for collaboration, research and media inquiries. It cannot be used to request emergency assistance.
People save lives. Information supports their decisions.
Back to top