TECHNOLOGY
TISC AI
We started developing AI technology for the purpose of implementing and utilizing Artificial Intelligence (AI) technology to respond to the 4th Industrial Revolution. Ultimately, we aims to actively participate in the establishment of an Artificial Intelligence eco-system and industrial development by establishing a demonstration system using an Artificial Intelligence platform, and among various ecosystems, we are developing technology to secure an intelligent integrated control platform for smart city creation and management.
In addition to the existing Camera-based video information for the integrated control platform, we aim to establish a control system using LiDAR Sensor. In particular, LiDAR is a three-dimensional point cloud data-based distance measurement method that can obtain precise results, and it is difficult to identify individuals because it is based on point cloud, not image or video, so we are building LiDAR Sensor-based control system.
Additionally, Edge computing, which can be efficiently replaced in the event of network problems or cloud paralysis, is emerging as a way to overcome cloud computing limitations. This enables big-data load balancing, and enables faster data collection and processing on the Edge to further address security concerns.
Smart City Intelligent Control Systems with all of the above technical objectives require the development of Real-time 3D data processing, multi-object recognition and segmentation technologies for Edge Computing Devices. And above all, we are developing technology to respond to the need for hardware acceleration technology for Artificial Intelligence lightening models.
Key Feature Developments
- Develop of FPGA DPU or Application processor GPU Based 3D Object Detection DNN Model
- Development of 3D LiDAR & Camera Sensor Based Object Detection / Object Tracking
- Development of Object (Car, Pedestrian, Cyclist, Etc.) Recognition Solution
- Install Multi-Sensor and Edge Device, Data Collection and Management
– Dataset Collection
– Dataset Labelling
– Dataset Learning One-Stop Service - Support for Real-time Inference Mode Based on AI Network
- Provide Wired and Wireless Telecommunication-based Cross-Platform 3D Visualization Tool to Support Live View / Sensor Setting / Dashboard Functionality
In addition to developing Artificial Intelligence technology, the ultimate goal of our Smart City Intelligent Control System is to connect devices using 5G technology to cope with the increase in the amount of mobile traffic and the increase in the number of objects. As a result, additional network infrastructure is being built, AI operating middleware running on Linux-based operating systems, architecture design for low power computing, and operation software development are continuously underway.
