Traffic congestion has become an increasingly concerning problem in modern society. Recent research has proven that Reinforcement Learning (RL) applied to Traffic Signal Control (TSC) is useful in mitigating congestion. In this paper, a model of real-world intersection with real traffic data collected in Hangzhou, China is simulated with different RL based traffic signal controllers. Two model free reinforcement learning methods are proposed namely: Deep Q-Learning (DQN) and double DQN (DDQN). These models are trained and tested at a 4-way intersection. Model adaptability and performance in different traffic scenarios are also measured and discussed in this paper.
Traffic jams at intersections increase congestion on the roads. Traditional traffic light systems control traffic for a fixed period, and traffic officers may also manually control traffic. Despite these established methods, intersection gridlock persists, prompting proponents to innovate traffic control systems. The adaptive traffic light network design used a combination of Artificial intelligence (AI) and IoT (IoT). The proponents used the MMDA's peak-hour traffic volume to develop fuzzy rules for the control system to define priority lanes and intersections using fuzzy logic. The historical data includes Taft Ave., T.M. Kalaw, Ayala Blvd., Finance Road, Romualdez St., Natividad Lopez, and Romualdez Street. Simulating these causes gridlock at all crossroads, causing delays and wait times. With the developed control system, the results are much more efficient, with a 50% improvement. The system's adaptability is also challenged, with increasing densities up to 80% indicating that the system is capable of handling the randomness in real-time zones.
Many autonomous systems (e.g., unmanned and autonomous vehicles) rely on computer vision algorithms to gather information about the surrounding environment and perform decision-making tasks accordingly such as tracking, detection, and segmentation. The performances of these systems can severally degenerate due to bad weather conditions such as rain and fog, which can result in erroneous and fatal decisions. Therefore, there is an urgent need to restore clear and rain-free background scenes from degraded videos in timely manner. In this paper, we propose a minimal-delay de-raining pipeline to eliminate noisy and unwanted rain artifacts from a live video feed. We propose a disentangled approach where we design a standalone deep learning Convolutional AutoEncoder (CAE) network to reconstruct noise-free images and then upgrade this architecture by adding a sequential Generative Adversarial Network (GAN) based module that enhances the quality of the image frames. Experimental results on rainstreak and raindrop datasets show that the proposed CAE architecture with GAN enhancing outperforms a benchmarking CNN for several image quality metrics. Other comparative simulations show that our proposed approach achieves close performances to an attentive standalone GAN de-raining approach with significant time complexity saving.
The “Last Mile Problem” is a global problem where commuters from the outskirts of the cities struggle to travel swiftly between their homes and required destinations. Recent research studies revealed that the ideal solution to this problem is ‘micromobility’ - small size, human or electric-powered vehicles used to travel shorter distances. This paper investigates the interest level of commuters towards micromobility, rental service preference, and high-demand geo locations in Colombo, the capital of Sri Lanka. From the literature review 28 variables were identified. A semi-structured survey circulated among commuters who frequently travel to the Colombo region, resulted in obtaining 132 responses. Based on the literature, four variables were identified for geospatial analysis. For this analysis, secondary data from a Colombo-based shared motorbike taxi service's trip record of 23,298 from 2019–20 were obtained. As per market segmentation and statistical analysis, two factors that would positively influence the interest levels of commuters were identified. A geospatial analysis conducted on secondary data revealed the high-demand geographical areas in Colombo.
Hyperloop is a cutting-edge high-speed rail transportation system. It can achieve aircraft-like speeds thanks to its unique configuration. Nevertheless, this very-high movement speed and the vacuum-tube environment result in multiple challenges to design the communication system. In this work, we propose a hybrid optical-wireless network architecture for Hyperloop communication system where data packets are transmitted from a centralized station through a backhaul optical link to multiple access points (APs) mounted on the tube. Each AP communicates with the moving pod using wireless signals. Then, we model the proposed architecture and analyze the downlink communication performance using queuing theory tools. We propose a design approach for the number of APs and the wireless technology to be implemented at each AP, considering several quality-of-service (QoS) requirements. We show, through multiple simulation examples, the impact of cell coverage and traffic intensity on the proposed design. It was demonstrated that the cost in terms of the required number of APs increases with strict probability of blocking constraints, whereas a wider bandwidth for the wireless link is required when the QoS constraint in terms of packet delay is alleviated.
