In order to avoid car accidents and ensure safer roads, Autonomous vehicles (AVs) have been created. These vehicles are capable of sensing its environment and navigating without human input, which is indeed a great step towards more secure experiences for drivers. As a method of improvement for the AVs, agent-based collision avoidance components that represent human cognition and emotions have been designed. However, agent-based solutions have not been validated using any key validation technique.
Considering this lack of validation practices, the authors F. Riaz and M. A. Niazi present in their upcoming paper state-of-the-art Emotion Enabled Cognitive Agent (EEC_Agent), which was proposed to avoid lateral collisions between semi-AVs. It is stated in the paper that the main drawback of EEC_Agent is that it is claiming utilization of emotions in making collision avoidance decisions, but no proper emotion generation mechanism has been proposed, which helps the cognitive agent to feel emotion according to the changing in the dynamic environment. In order to overcome this problem, they have redesigned the architecture of EEC_Agent using EABM (Exploratory Agent Based Modeling) and explored the role of OCC (Ortony, Clore & Collins) model in the fear generation of EEC_Agent.Emotion Enabled Cognitive Agent Architecture
The authors utilize a SimConnector approach to join fuzzy logic environment results with NetLogo agent-based simulation. VOMAS (Virtual Overlay MultiAgent System) Agent is used to validate the performance of EEEC_Agent (Enhanced Emotion Enabled Cognitive Agent) in the rear end and overtaking collisions. The extensive experiments presented in the paper proved that validated EEEC_Agent can perform collision avoidance with smaller SSD (Stopping Sight Distance) and OSD (Overtaking Sight Distance) as compared to the human driver.