Six of Singapore’s 17 reservoirs will soon be monitored by Beyond Visual Line of Sight (BVLOS) unmanned drones, PUB, Singapore’s national water agency, announced.
PUB plans to monitor the reservoirs with drones to save about 5,000 man-hours. Mr Yeo Keng Soon, director of PUB’s Catchment and Waterways Department, said that taking care of all 17 reservoirs is a challenge manpower-wise to monitor, in addition to ensuring the reservoir’s condition. He also said that the use of drones is in line with the Board’s commitment of “leveraging technology as part of the SMART HUB roadmap” to improve operations and meet future needs.
Normally, 7,200 man-hours are spent in a year conducting daily patrols to check for excessive aquatic plant growth and algal blooms. Data on water activities like fishing and paddling activities are also collected during these patrols to ensure that these activities are being done safely. With the use of drones, the Board will be able to survey large areas of the reservoir, collect comprehensive data and trigger alerts through the drones when illegal or suspicious activities are detected.
ST Engineering will provide the drones and DroNet, the company’s proprietary drone operating system, as part of its contract with PUB. DroNet already passed PUB’s trials last year and has been customised to meet PUB’s requirements.
One drone will be deployed at each of the six reservoirs and will follow a pre-programmed flight path within the compound. A nearby operator will monitor the drone as it flies through the reservoir. MacRitchie and Marina Reservoir will be the first to have monitoring drones, while Serangoon, Kranji Lower Seletar and Lower Peirce will receive theirs later in 2021.
As mentioned before, the drones are capable of monitoring water quality, aquatic plant growth and water activities. These are possible through the many systems and cameras that are onboard the drone. First, the drone’s water quality remote sensing system analyses water turbidity and algae concentration to acquire a good correlation to actual water quality. Meanwhile, a video analytics algorithm will use video data collected during earlier trials to identify aquatic plant overgrowth in the reservoir through the drone camera’s live video feed. Lastly, the drone camera’s live video feed will capture water activities while a video analytics algorithm flags potential concerns (e.g. fishing in non-designated areas or overcrowding of vessels in a particular area.)
An online dashboard will display to PUB officers the statistical data and live video feed each drone transmits. Meanwhile, near real-time alerts on illegal water activities will be sent to a dedicated Telegram channel that officers can access with their mobile phones. They will then be able to prioritise urgent cases that may pose a potential danger to the public and respond on time.
Written by John Paul Joaquin