Pola Medical Releases 2026 Updated Version of Heatstroke Prevention AI Camera "Kaokara"
Pola Medical, part of the Pola Orbis Group, has launched the 2026 updated version of its heatstroke prevention AI camera "Kaokara" to tackle severe heat issues in workplaces. This update significantly enhances heatstroke risk assessment accuracy and detection robustness, making it resilient to lighting variations and facial obstructions. It also introduces improved operational features like individual face image sorting and refined email notifications.
📋 Article Processing Timeline
- 📰 Published: April 2, 2026 at 20:10
- 🔍 Collected: April 2, 2026 at 13:36
- 🤖 AI Analyzed: April 16, 2026 at 01:18 (323h 41m after Collected)
Pola Orbis Group's Pola Medical Inc. (Headquarters: Yokohama, Kanagawa Prefecture, Representative Director and President: Tsuyoshi Matsumoto) will release the 2026 updated version of its heatstroke prevention AI camera "Kaokara" to address serious heat-related issues in workplaces.
This update improves heatstroke risk assessment accuracy and achieves robust detection performance that is less affected by lighting variations (partial differences in brightness and darkness) due to shooting environments, and facial obstructions such as masks and sunglasses. Furthermore, various functional improvements have been made in terms of operation, such as a sorting function for individual face images. Through these improvements, we will further promote advanced initiatives in on-site safety management.
Update Details
1. Improved Judgment Accuracy and Robustness
Based on approximately 2 million face image data collected in fiscal year 2025, the AI model has been redesigned to improve the detection accuracy of facial color, expression, and perspiration. In addition, by re-learning using images acquired under various human and environmental conditions, the certainty of prediction has improved by more than 10% even for images with quality issues such as uneven lighting. In particular, it has been confirmed that the improvement effect is greater for images with severe shooting conditions.
2. Sorting of Individual Face Images (Beta Feature)
A new function has been implemented to narrow down judgment results for each individual based on the similarity of face images. This allows administrators to view a list of judgment results for specific individuals.
3. Improved Email Notification Function
The function has been improved to allow selection of alerts to be notified via email. This makes it possible to notify administrators in real-time only of the necessary judgment results according to on-site operations.
This update will be automatically applied after 2:00 PM on April 3rd by connecting the product online. For details on functions and consultations regarding introduction, please feel free to contact the following window.
Kaokara Homepage
https://kaokara.jp/
POLA Medical Kaokara Customer Support
TEL: 0120-454-025 (Weekdays 8:30 AM - 6:00 PM)
Email: poh_polamedical_support@po-holdings.co.jp
This update improves heatstroke risk assessment accuracy and achieves robust detection performance that is less affected by lighting variations (partial differences in brightness and darkness) due to shooting environments, and facial obstructions such as masks and sunglasses. Furthermore, various functional improvements have been made in terms of operation, such as a sorting function for individual face images. Through these improvements, we will further promote advanced initiatives in on-site safety management.
Update Details
1. Improved Judgment Accuracy and Robustness
Based on approximately 2 million face image data collected in fiscal year 2025, the AI model has been redesigned to improve the detection accuracy of facial color, expression, and perspiration. In addition, by re-learning using images acquired under various human and environmental conditions, the certainty of prediction has improved by more than 10% even for images with quality issues such as uneven lighting. In particular, it has been confirmed that the improvement effect is greater for images with severe shooting conditions.
2. Sorting of Individual Face Images (Beta Feature)
A new function has been implemented to narrow down judgment results for each individual based on the similarity of face images. This allows administrators to view a list of judgment results for specific individuals.
3. Improved Email Notification Function
The function has been improved to allow selection of alerts to be notified via email. This makes it possible to notify administrators in real-time only of the necessary judgment results according to on-site operations.
This update will be automatically applied after 2:00 PM on April 3rd by connecting the product online. For details on functions and consultations regarding introduction, please feel free to contact the following window.
Kaokara Homepage
https://kaokara.jp/
POLA Medical Kaokara Customer Support
TEL: 0120-454-025 (Weekdays 8:30 AM - 6:00 PM)
Email: poh_polamedical_support@po-holdings.co.jp
FAQ
What are the main improvements in the 2026 updated version of "Kaokara"?
The main improvements include enhanced heatstroke risk assessment accuracy, robust detection performance less affected by lighting variations and facial obstructions, a beta feature for sorting individual face images, and improved email notification functions.