Camille Kamga

Bahman Moghimi, Navid Kalantari, Camille Kamga et al.

Capacity expansions as well as its reduction have been widely anticipated as important countermeasures for traffic congestion. Although capacity expansion had been traditionally well noticed as a congestion mitigation measure, but it was not until recently that capacity reduction measures such as; congestion pricing, road diet and other such capacity reduction measures were noticed as congestion mitigation measures. Measures such as signal optimization, metering and congestion pricing are intended to affect the travel pattern and assignment behavior of travelers to make the results of the User Equilibrium (UE) traffic assignment, followed by the travelers, closer to the System Optimal (SO) outcomes, intended by the planners. As such, a bi-level optimization model was formulated for the simultaneous optimization of capacity improvement/expansion and signal timing in an urban transportation network. The model takes into account both the effect of higher demand, induced by the capacity expansion, and the rerouting potential effect of traffic signals. The solution algorithm developed here consists of two major components; the gradient projection algorithm (GP) to solve the lower level traffic assignment problem and the Simulated Annealing (SA) algorithm to solve the master problem. In order to illustrate the method a case study was examined. The analysis and application of the proposed algorithm shows that the performance function gradually converged along the simulation run, and no divergence problem is observed. By applying the developed algorithm, the total network travel time reduced by 13.42%, in which 5.76% is reached by only optimizing the signal times. Using the GP algorithm and taking advantages of dynamic memory in the process of simulation, the whole simulation acquired from computer time is 13.63 seconds with a convergence rate of 0.001.

Yiqiao Li, Jie Wei, Camille Kamga

Heavy-duty trucks pose significant safety challenges due to their large size and limited maneuverability compared to passenger vehicles. A deeper understanding of truck characteristics is essential for enhancing the safety perspective of cooperative autonomous driving. Traditional LiDAR-based truck classification methods rely on extensive manual annotations, which makes them labor-intensive and costly. The rapid advancement of large language models (LLMs) trained on massive datasets presents an opportunity to leverage their few-shot learning capabilities for truck classification. However, existing vision-language models (VLMs) are primarily trained on image datasets, which makes it challenging to directly process point cloud data. This study introduces a novel framework that integrates roadside LiDAR point cloud data with VLMs to facilitate efficient and accurate truck classification, which supports cooperative and safe driving environments. This study introduces three key innovations: (1) leveraging real-world LiDAR datasets for model development, (2) designing a preprocessing pipeline to adapt point cloud data for VLM input, including point cloud registration for dense 3D rendering and mathematical morphological techniques to enhance feature representation, and (3) utilizing in-context learning with few-shot prompting to enable vehicle classification with minimally labeled training data. Experimental results demonstrate encouraging performance of this method and present its potential to reduce annotation efforts while improving classification accuracy.