Functional Risk for AV using AI Hybrid Deterministic Method

Avatar for Dan CHIA
Dan CHIA    
Associate Professor

Read More 

Avatar for Yiyang PEI
Yiyang PEI    
Associate Professor

Read More 

Avatar for Peter WASZECKI
Peter WASZECKI    
Assistant Professor

Read More 

Avatar for Jianxin ZHENG
Jianxin ZHENG    
Associate Professor

Read More 

Avatar for Cindy Goh (UOG)
Cindy GOH (UOG)    
Researcher
Avatar for Sye Loong Keoh (UOG)
Sye Loong KEOH (UOG)    
Researcher
Avatar for Chris Johnson (UOG)
Chris JOHNSON (UOG)    
Researcher
Avatar for Dilip Limbu (MOOVITA)
Dilip LIMBU (MOOVITA)    
Researcher
Avatar for Anthony Wong (MOOVITA)
Anthony WONG (MOOVITA)    
Researcher

In this project, a hybrid deterministic method is used for risk assessment within an ADS. 

Problem statement

One of the key challenges in deploying Autonomous Driving Systems (ADS) on public roads is ensuring safety. This includes exploring how existing technologies, such as AI and hybrid methods, can enhance risk assessment for ADS.

Solution and Notable Contribution:

By detecting environmental anomalies, the system triggers safety actions to enhance overall safety. A MoCAS (Mobile Camera Acquisition System) was developed to perform these assessments, leading to a larger external grant. The results were subsequently licensed to our collaborator, MooVita.
 

Surveillance camera footage showing a parking area with an automated pedestrian detection system. The system overlays colored grids (green, yellow, and red) over the road and pedestrian paths to monitor zones, and draws bounding boxes around people to calculate risk factors in real-time.

 

A technical diagram and a corresponding photograph of a custom mobile surveillance unit, model WP-806. The schematic shows the unit's dimensions and front and back views, while the photograph shows the assembled white metal stand equipped with a camera, monitor, and equipment cabinet on casters.

 

Through active engagement with stakeholders, development of user-friendly applications, gamification strategies, and adaptable technical infrastructure, LEARN positions itself as a comprehensive and dynamic solution to advance MTL education. By leveraging the collaborative efforts of teachers, schools, and students/parents, LEARN is poised to make a lasting impact on the language learning landscape.