Mardhani Riasetiawan

Road Lane Segmentation Using Deconvolutional Neural Network

Lane departure warning (LDW) system attached to modern vehicles is responsible for lowering car accident caused by inappropriate lane changing behaviour. However the success of LDW system depends on how well it define and segment the drivable ego lane. As the development of deep learning methods, the expensive light detection and ranging (LIDAR) guided system is now replaced by analysis of digital images captured by low-cost camera. Numerous method has been applied to address this problem. However, most approach only focusing on achieving segmentation accuracy, while in the real implementation of LDW, computational time is also an importance metric. This research focuses on utilizing deconvolutional neural network to generate accurate road lane segmentation in a realtime fashion. Feature maps from the input image is learned to form a representation. The use of convolution and pooling layer to build the feature map resulting in spatially smaller feature map. Deconvolution and unpooling layer then applied to the feature map to reconstruct it back to its input size. The method used in this research resulting a 98.38% pixel level accuracy and able to predict a single input frame in 28 ms, enabling realtime prediction which is essential for a LDW system.

Payload Formats and Universal Data for Near Field Communication

NFC has a variant where Type 1 with NFC-A, Type 2 with NFC-A, Type 3 with NFC F, and Type 4 with NFC-B. In its application, the use of NFC is based on its tag type. If we go further, each type has similarities and differences. We can look at the various parameters, but for now we will look at 4 parameters, namely Sequence format, Bit level coding, frame format and data & payload format. From the four parameters, compared to the parameter value and obtained the difference and the equation. With that use, an mobile application only runs in one type, while for another type must be modified. To overcome this, it is proposed a framework that can handle the tags of different types. Proposals are made for universal data and payload formats, and the expectations can be recognized by all NFC tag reading tools. Architectures are built with attention to the similarities and differences for each type. For above four parameters. Beginning with getting the data for the use of the most tag type, then made the proposed universal data format. Commonly used types are types 1 and 2. Of the two types, the differences of the four parameters are on the data and payload parameters. And more specifically on the CRC. Thus, a universal data and payload format with a specific CRC is used. General Terms Embedded system, proximity card technology.

Cloud-Based Processing on Data Science for Visualization

The big data processing and visualization have the challenge on method and process. The volume, variety, velocity, and veracity in the big data need to handle for visualizing the data. The research work investigates, design and develop cloud-based processing for data science visualization. The cloud-based processing used the data management that interact with Google drive, communicate with processing tools that deliver the system, and visualize with Google Fusion. The research uses financial banking data in Indonesia, that analyze based on geolocation, transaction flows, and networks. The research process consists of data mapping process, data tagging, data manipulation, and data visualization to resulting the data science visualization. The cloud-based processing present the data science process can be done use cloud. The research has a contribution on delivering cloud-based approach to handling data science visualization for big data.

Data Center workload performance in dynamic resources allocation

The resource allocation is an essential aspect of Data Center performance. The strategy between dynamic and dedicated resource allocation have objectives in performance and reliability. The performance consists of workload, usage, and processing performance. The study works in workload performance test in dynamic resource allocation. The research present the correlation between load handling memory usage impacted the performance index. The relation between the aspect gives the behavior and characteristics in the workload performance. The result can be used for a baseline to identify and manage the resource management and reach the reliable performance of Data Center.

Distributed Replicated Block Device (DRDB) implementation on cluster storage data migration

Data Center systems are required to have high availability in order to meet the continuing needs of the user. If the server or application in which a failure or require maintenance, the virtual machine will be migrated to another server in the cluster are still available. The role of shared storage is very important here to keep a virtual machine that can continue the work and do not lose data on the destination server. This study seeks to implement virtualization at the server cluster system uses a virtual machine on Proxmox VE and Distributed Replicated Block Device (DRBD) as shared storage. Implementation is done by using two nodes, and made a comparison with two other nodes in the cluster system that does not use shared storage. Shared storage works by synchronizing and replication of virtual machine data that can then migrate online. The use of shared storage will affect virtual machine performance, especially on the speed of the disk during the process of sending and receiving data, and the availability of services. Measurement of downtime during the migration to test the success of the system. Testing the TCP connection is made to ensure the network throughput and connection test results compared to test disk performance.