View Proposal

---

Proposer

Kah Kit Ng

Title

Mitigating Phantom Traffic Jams through Real-Time Behavioral Feedback: A Human-in-the-Loop 2D Simulation Study

Goal

1. To engineer an interactive, top-down 2D highway simulator using Python (Pygame) that accurately models continuous vehicle kinematics (acceleration, reactive braking, and engine-drag coasting). 2. To program an AI vehicle platoon utilizing the Intelligent Driver Model (IDM) that reacts realistically to the human player, visually demonstrating the formation of a backward-propagating traffic shockwave. 3. To develop a Rule-Based Telemetry Engine that silently logs player kinematics (velocity, lane position) and triggers real-time visual UI alerts when the player induces a shockwave or lane-hogs. 4. To conduct a controlled Human-in-the-Loop behavioral experiment comparing a control group (no feedback) against an experimental group (real-time UI feedback) to quantify the reduction in aggressive braking events.

Description

Malaysian highways, particularly within the Klang Valley, frequently suffer from "phantom traffic jams"—kinematic shockwaves that ripple backward through traffic platoons without any physical obstruction. Foundational traffic flow research confirms that these shockwaves are directly triggered by human error, specifically delayed reaction times followed by harsh, reactive braking. In behavioral economics, driving represents a classic network externality: drivers look out the windshield and rarely perceive the cascading macroscopic consequences they leave in their rearview mirrors. While automotive engineering provides solutions like Deceleration Fuel Cut-Off (DFCO)—allowing vehicles to coast without burning fuel or using brake pads—many drivers default to "panic braking" or "lane hogging" due to a lack of situational awareness. This project proposes a Human-in-the-Loop (HITL) 2D driving simulation to test whether real-time, in-cabin visual feedback can successfully rewire these detrimental habits and promote anticipatory coasting.

Resources

https://www.nhtsa.gov/sites/nhtsa.gov/files/distraction_npfg-02162012.pdf

Background

Python Programming; Pygame; Pymunk;

Url

Difficulty Level

High

Ethical Approval

Full

Number Of Students

1

Supervisor

Kah Kit Ng

Keywords

microscopic simulation, behavioural conditioning, human computer interaction (hci)

Degrees

Bachelor of Science in Computing Science