JDSC Selected for "Hiroshima AI Sandbox" Phase 2 to Develop Digital Twin for Livestock Industry Tackling Climate Change, Joint Demonstration with Hiroshima University

JDSC Inc. (Headquarters: Bunkyo-ku, Tokyo; Representatives: Satoshi KATO ELTES / Asuka SATO; hereinafter "JDSC") has been selected as a recipient for the second phase of the "Hiroshima AI Sandbox" program, implemented by Hiroshima Prefecture. In collaboration with the Dairy Ecosystem Technology Development Center at the National University Corporation, Hiroshima University, JDSC will work on the development and demonstration of a digital twin application for the cattle shed environment.

*1 This is the first commercial digital twin implementation in the livestock industry in Japan, according to our research.

<Project Overview>

Content: Implementation of a digital twin in the livestock industry and construction of an optimization application for animal husbandry – A mechanism to visualize the "invisible heat" in cattle sheds.

Matching Partner: National University Corporation, Hiroshima University, Dairy Ecosystem Technology Development Center

Demonstration Period: June 2026 – February 2027 (approx. 8 months)

Demonstration Field: Hiroshima University Farm (scale of 30 milking cows)

<Background>

With the acceleration of climate change, the importance of heat stress countermeasures in dairy farming is rapidly increasing. It is reported that when the THI (Temperature-Humidity Index)*2 exceeds 72, dairy cows can experience a milk yield decrease of approximately 2kg per cow per day*3. However, in reality, "invisible heat" is often overlooked, and the effectiveness of facility investments cannot be verified.

*2 THI = Temperature-Humidity Index. An international standard index for evaluating heat stress in livestock.

*3 West, J.W. (2003)

<Solution Features>

This project will visualize the temperature and humidity environment inside cattle sheds in real-time using 3D, enabling an understanding of where cows are experiencing heat stress on a "surface" level. Furthermore, by combining causal inference and machine learning, it will quantitatively analyze "which facility improvements will lead to milk yield recovery." Features such as natural language advice generated by generative AI and a What-If simulation function to pre-calculate the ROI of facility investments will be implemented. Through these, we aim to recover 50-75% of economic losses caused by heat.

<About the Collaboration Partner>

The collaborating partner, the Dairy Ecosystem Technology Development Center at Hiroshima University, is a research hub established in 2025, possessing the only university-affiliated dairy department in the Chugoku, Shikoku, and Kyushu regions. Equipped with leading research facilities in Japan, including milking robots, calf feeding robots, individual animal management systems, and next-generation cattle shed environment systems, it is advancing cutting-edge dairy AI and data research, such as cattle weight estimation using LiDAR x machine learning, rapid mastitis diagnostic technology, and THI prediction AI models.

JDSC and Hiroshima University have prior experience in joint proposals within the context of smart livestock and smart agriculture. In this project, they will collaborate closely on providing the demonstration field, sharing on-site needs, data provision, and AI/UI evaluation.

<Future Development>

This project aims to establish Hiroshima University as a leading domestic research and demonstration base and to preferentially deploy solutions from dairy farmers in Hiroshima Prefecture. Through collaboration with the Hiroshima Prefectural Dairy Association and JA Hiroshima, a phased rollout from the prefecture to the Chugoku, Shikoku, and Kyushu regions, and then nationwide, is planned. The ultimate target is approximately 5,400 dairy farms nationwide (with 50 or more adult cattle)*4.

After commercialization, a recurring model of Digital Twin as a Service (DTaaS) will be adopted. For dairy farms with 80 head, an investment recovery period of approximately 1.1 years is projected*5. Through direct income improvement by reducing heat loss by 50-75%, it will contribute to preventing farm closures and enhancing the sustainability of regional dairy farming.

*4 According to the 2025 Livestock Statistics.

*5 Calculation for dairy farms with 80 head. Assumes annual heat loss of \1,440,000 (calculated based on milk yield decrease when THI exceeds 72) and labor costs of \12,400,000 per year (\155,000 per head), with an introduction effect of 70% heat loss avoidance and 10% labor cost improvement. System introduction cost is estimated at \1,000,000, annual basic usage fee at \360,000, and annual per-head charge at \192,000. Costs for facility management improvements are not included due to difficulty in calculation.

Comment from Hiroyuki Taguchi, Executive Officer at JDSC:

"Heat stress due to climate change is having a serious impact on Japanese dairy farming operations. This project will visualize the effects of heat, which have been considered 'invisible' until now, through a digital twin of the cattle shed environment and causal inference AI, thereby building a system to support dairy farmers' decision-making. It is highly significant to be able to collaborate with Hiroshima University, which possesses the only university-affiliated dairy facility in the Chugoku, Shikoku, and Kyushu regions, and we will deliver this model case from Hiroshima to dairy farmers nationwide."

About "Hiroshima AI Sandbox" https://hiroshima-ai-sandbox.jp/

This initiative supports the development of AI solutions by matching AI developers nationwide with companies and local governments in Hiroshima Prefecture to solve industrial and regional issues within the prefecture. It aims to foster new innovation ecosystems by attracting companies and talent to Hiroshima who are challenging new businesses using AI.

About JDSC Inc. https://jdsc.ai/

JDSC is a technology company driving the transformation of Japan's core industries, centered on AI agents and data science. By deeply engaging with sites in manufacturing, logistics, infrastructure, and public sectors, it achieves everything from advanced decision-making through DX and AX to the optimization and autonomy of the real world via physical AI. It contributes to upgrading Japan by creating new industrial standards that solve social issues facing societal challenges.