Service Needs

Reference No.

N-0184

Title

AI Environmental Air Nuisance Investigation Robot Dog

Business Needs / Challenges

Every year, the Environmental Protection Department (EPD) receive over 1 200 complaints against environmental air nuisances, most come from unclear sources, and we conduct about 3 000 relevant site investigations annually. Thousands of affected people may experience discomfort, nausea or even dizziness. In some cases, prompt investigation is vital in finding out the source of pollution in order to quickly stop the nuisance, in particular for cases involving the vulnerable children.

Conventionally, EPD investigators had to find out the nuisance by smelling and approach the direction of the emission source. Safety and health concerns for frontline colleagues were raised inevitably from prolonged investigations, especially for environmental nuisances of Mysterious Gas Incident (MGI). While gaseous detectors would be used to measure the concentration of pollutants, they could only provide limited information such as concentration. The investigation was still highly dependent on human searching.

The introduction and development of AI Environmental Air Nuisance Investigation Robot Dog (AI-Dog) would turn the tile. It would be equipped with computer vision, robot mechanics, advanced gaseous sensors and AI algorithms to predict the source and pollution activity, which would track the location of the source autonomously with tailor-designed directional air intakes. Designing to have 5G linkage with IoTs such as the computer servers of Control Centre of EPD, it can also access big data for air emission inventories of nearby inspection area. Through continuous AI machine learning, its accuracy and effectiveness would be improved over time.

AI-Dog would break the new ground in environmental monitoring and emergency response. By leveraging advanced AI technologies, this autonomous robotics system can effectively identify pollution sources and locate them. Its autonomous nature eliminates the need for human involvement in unhealthy or unsafe areas during the investigations, making it a valuable companion for government investigators and citizens.

Application Areas

Environment
Health

Expected Outcomes

To complete the Proof-of-Concept (PoC) development within six months, and provide the required equipment and AI training data as necessary.

Technologies to be Used

Artificial Intelligence (AI)
Data Analytics
Internet of Things (IoT)
Machine Learning
Robotics