SoundEye

SoundEye provides advance sound recognition solutions capable of classifying sounds. Our solution is currently deployed in airport, hospital, shopping mall, condominium, research labs, server room, and nursing homes. Our products include a smart recessed light and a smart surveillance system that can detect sounds, including screaming, gunshot, coughing, explosion, etc. We have currently 6 staff. SoundEye is a social enterprise and our vision is to see the world through sound – by doing so, improve productivity and save lives! Though we are a social enterprise we are still generating healthy revenue – total cash sales received from June 2020 till Dec 2020 (current fiscal year = June 2020 till May 2021) is over S$524,000 (signed purchase order = S$835,000).

PUCK is a modular system that is capable of performing sound recognition and allow the interfacing of other sensors (motion, laser, temperature, smoke, etc.) to perform various uses-cases. One of the most sought the application is to detect occupancy inside washroom cubicle and handicap washroom. For emergency monitoring purpose our solution is able to trigger alert upon detection of a person screaming or not moving for a period of time. SoundEye focus on B2B and PUCK was deployed to the elderly nursing home and social service organizations (orphanage, home for destitute, institute for mental impairments, etc.)

235,000 people over age 15 visit emergency rooms because of injuries suffered in the bathroom. SoundEye at this stage is able to detect a person screaming for help and detect a person not moving inside the washroom for a period of time.

For this proposed R&D project, we are hoping to develop a system that can predict and detect a fall in the washroom using sensors that do not intrude privacy. To predict fall, assessment using the sensor to evaluate how a person is trying to stand up from the toilet bowl will be executed unobtrusively. This assessment does not require wearable sensors nor a therapist to conduct. The same sensor can also detect a person on the floor, thus achieving fall detection.

The CTO and lead software engineer of SoundEye have experience in vision analytic. SoundEye will be developing the analytic module capable of predicting and detecting a fall inside a washroom using the depth maps produced by the sensor. The partner from Israel will need to provide a sensor that is capable of detecting human motion only.

SoundEye will also be responsible for the testing and deploying the proposed solution to the real world with stakeholders in Singapore.

The partner to provide a sensor capable of detecting the fine-motion produced by the human. The solution must not use any sensor, such as a camera that will intrude the privacy of the human. The dimension of the sensor should be small to be deployed for indoor and on the ceiling. The design should take into consideration ease-of-deployment and affordability. Last but not least, should the sensor need to perform any process/analytics, it should be done on the device directly (edge computing).

The sensor must provide accurate data and must not be affected by the humidity, temperature and other conditions that you may find inside a conventional washroom, or any other indoor environment. The depth mapping captured should observe human only while filtering out the environment.