Earticle

As autonomous driving and connected car technology advance, various deep learning applications for autonomous vehicles and complex traffic situations are increasing. Autonomous vehicles must collect and process vast amounts of sensor data to support various deep learning applications, but vehicles have limited computing resources to perform complex deep learning operations. Therefore, edge computing is a promising solution to complement the limitations of autonomous vehicles. In this paper, we design edge computing for efficient task processing in an autonomous driving environment using a driving simulator. Also, we propose a task scheduling and offloading method which determines the target server to offload a task according to the characteristics of the task and the computing resources. The effectiveness of the proposed method is verified through experimental evaluation in an autonomous driving environment, supporting multiple deep learning services that we established by using a driving simulator.

This paper proposes a motion planning method for connected autonomous vehicles approaching an intersection where there is no traffic light. Crossing an intersection poses a new set of challenges due to the various interaction with opposing vehicles. By exploiting the position, velocity, heading, and trajectory of vehicles through V2X communication, the future motion of vehicles in the vicinity of the intersection can be predicted. The predicted motion is translated into an SLcoordinate frame where the likelihood of a path conflict can be predetermined. This information is fed into the motion planner, where it calculates the optimal trajectory of the vehicle using a search-based method. The proposed motion planner is validated using two identical CAVs in a no-traffic-light within the Incheon Songdo Technopark testbed.

In the present scenario, the self-driving industry is in the commercialization stage of the consumer market. There are no longer any difficulties simply detecting an object and performing certain tasks in terms of vehicle control and navigation. Nevertheless, unexpected problems arise due to occlusion, unbalanced ground, and bad weather. This paper presents a method for correcting inaccurate distance recognition results when a vehicle vibrates against the unbalanced ground. The proposed method follows a pipeline of preprocessing techniques such as LiDAR-camera calibration, 3D point cloud data acquisition from LiDAR and camera to model 3D plane equations in RANSAC. The distance error is corrected by estimating the orientations from an IMU sensor in real-time and by assessing Rotation and translation in accordance with the pose changes of the vehicle.

In this paper, we propose a dead reckoning method for estimating the current position using geographic information including curvature, inclination, and azimuth. We describe how we match the posture, position, and speed with the IMU measurements and suggest how we can correct the location with the help of the geographical information. The proposed dead reckoning method along with the geographic information, will be helpful for more accurate position estimate because the user stays along the path easily identified by the geographical information in most applications.

This paper introduces an efficient non-binary low-density parity-check (NB-LDPC) decoder architecture, in terms of increasing decoding throughput. By taking advantages of non-binary quasi-cyclic LDPC codes, a new layered decoding algorithm and corresponding efficient hardware architecture are introduced. The proposed method can improve parallelism in decoding estimations of NB-LDPC decoder while remaining error-correcting performance. The implementations results confirmed that the proposed decoder with two threads can achieve a throughput of about 2.78 Gbps, which is around 1.63 times faster than that of the state-of-the-art decoder at almost the same hardware efficiency.

This paper presents the design process of an Eddy Current Brake (ECB) system for an electric scooter application. In order to obtain sufficient braking force in the limited space of the electric scooter’s braking system, the impact of three key variables, air gap length, brake disc thickness, and magnet grade, are studied and analyzed by using three-dimensional (3D) finite element (FE) analysis. The braking torque of the ECB is also analyzed using FE analysis. Finally, an experimental setup is fabricated using a 1 kW brushless DC (BLDC) motor that was used in a commercial electric scooter, and preliminary results of measuring the rotational speed of the brake disc are presented.

This paper presents a new method applied to photoplethysmography (PPG) signal with compression technique that is effectively applied to wearable devices. This data compression technique is especially effective for applications mainly focused on the user's heart rate. We applied this compression technique to PPG signals with sampling frequencies of 125 Hz (BIDMC Physionet). This method has the highest compression ratio (CR) as 77.8. Besides, the new method applies effectively to heart rate estimation without requiring complicated calculations. The proposed method can be applied to real-time system with PPG signal measurement with low-speed microcontroller unit (MCU).

Fires are one of the most frequent natural disasters in the world. To minimize the damage from the fire, it is important to build surveillance systems for detecting fire with fire detection algorithms, and a variety of deep learning networks have been proposed with sufficiently high accuracy. However, regardless of the performance of the existing solutions, such systems are bound to a fixed environment so that its coverage is limited. Therefore, in this paper, we propose a fire detection system using mobile robots so as to overcome this limitation. Based on a freely moving robot driving system, the proposed system consist of IoT devices, edge cloud computing, and fire detection algorithm. We hope our proposed system offers useful insights to building practical mobile fire detection systems for minimizing the damages caused by fire.