About the Client

The client is an AI-driven healthcare solution committed to innovate and support the global community for patient healthcare and to address the COVID-19 global pandemic by utilizing technology and portability.

Their already existing solution allows uninterrupted continuity of care for patients in different settings, from acute hospitals to rehabilitation facilities and home care.

Thus, reducing the incidence of complications and amputations, and leading to faster recovery and significant cost savings for care providers and to the industry. 

They have partnered with Integrated Health Information Systems (IHiS) (technology agency for public healthcare) to develop a temperature screening solution powered by Artificial Intelligence (AI) that screens and identifies febrile persons.

About the Project

Temperature screening is one of the most common ways of detecting initial symptoms of viral infections, such as required during the current COVID-19 pandemic.

However, temperature screening with forehead thermometers requires proximity, is time-consuming, manpower intensive and creates long queues that could increase the chance of spreading infection.

As forehead thermometers are not meant for mass screening, their rate of malfunction and accuracy in identifying febrile cases may not meet the requirements of mass screening.

This solution can significantly reduce the workload of temperature screening staff.

Instead of taking forehead temperatures and needing close contact with every person, screening staff only need to manually record the temperatures of suspected cases detected by the application.

The solution performs the mass screening and alerts the staff with audible sounds and visual prompts identifying the febrile person, while staff can perform other tasks.

When alerted to a febrile person, staff can then confirm with forehead or ear (tympanic) thermometers.

This process can result in significant productivity gains, avoidance of human error and smoother flow of people into facilities while ensuring safety.

The anonymous, non-personal data collected from screenings (location, device, time, temperature and alert) can be sent to a dashboard to support operations monitoring.

The components in this platform are:

  • SmartPhone/Tablet Device for Temperature Screening
  • Web Portal for Application Administrator

The application was developed based on Enterprise Singapore’s fever guidelines, specifically “Singapore Standard 582 – Specification for thermal imagers for human temperature screening.”

Rigorous tests and trials were carried out on individuals, hospitals, polyclinics, commercial buildings and dormitories to ensure best-in-class temperature screening for both Type 1 (walk past tests) and Type 2 (standing test).


The platform is a lightweight solution that uses a smartphone/tablet fitted with thermal and 3D laser cameras.

It is currently one of only a few temperature screening devices in Singapore that has been registered as a Class B medical device for general hospital use by the Health Sciences Authority.

The AI application processes and analyses the images from the smartphone/tablet camera (which captures facial features) and maps them to images from the thermal camera (which measures temperatures) as well as the laser camera (which measures distance). This Smart Health solution powered by AI can:

  • Recognize human facial features from thermal images to measure the forehead temperature, including individuals wearing spectacles, surgical masks, hats and other headgear, even while they are walking. The solution is able to handle an estimated 5,000 individuals a day.
  • Calibrate the temperature reading taking into consideration the ambient temperature.
  • Incorporate distance analysis and compensation (as temperature measurements reduce with distance). It measures temperatures accurately up to a distance of three meters from the camera – a feature not found in other similar range solutions.
  • Provide audio alerts to indicate that a temperature has been recorded, as well as trigger alarms for secondary checks and send SMS/email notifications to flag febrile persons to multiple users.
  • Provide status reports and real-time updates via the dashboard with rate of traffic and the number of febrile persons identified. The dashboard can be connected to multiple cameras at different locations and enable remote monitoring of the status at different sites.
  • Screening made easy: Screen a group of eight to 10 people at one time
  • No more long ques – There is no need to stop in front of the application connected device. Just walk on by 
  • Fuss free detection – An audio alert will sound if a person shows signs of having fever. The prompts will enable secondary check for a manual temperature reading
  • Flexible deployment – The software calibrates temperature screening by taking into consideration the ambient temperature. It takes up a small amount of space and can be easily redeployed
  • Remote monitoring across various sites – Get real time updates via dashboard with rate of traffic and the number of persons who might have a fever.

Device Setup

Step 1 – The device is set up at the screening point and calibrated to consider the ambient temperature.

Step 2 – Use crowd barriers to funnel your crowd to a narrow point

Step 3 – Let them walk on by, but pay attention to the beeps of the device

Step 4 – If the device sounds an alarm, it means the person might be having a fever. Pull them aside to check their temperature with a handled thermometer. 

Step 5 – Do random manual checks with handheld thermometers.



Technology Stack Used

Module Recommended Technology Stack
Android Kotlin
Mobile SDK Google ML Kit
Backend C++
ML Software Library TensorFlow
Thermographic Sensing FLIR
Real-time Computer Vision OpenCV

Looking for a similar App ?