Research on Autonomous Electric Vehicle Developed in LIPI


A. Telecommunications System for Autonomous Electric Vehicles

We are developing a telecommunications system for autonomous electric vehicles which will be operated in limited environments such as office complexes, industry, plantations, and independent city models. The autonomous electric vehicle will have technical characteristics in carrying a limited number of passengers (one to two passengers). In addition, the vehicle speed tends to be slow, and the energy consumption is relatively small compared to autonomous electric vehicles developed abroad. The development of this autonomous electric vehicle is chosen because it considers the availability of battery technology and the differentiation of similar vehicles designed by large corporations and developed countries. And also to simplify the research and development process by reducing complex licensing and standardization aspects as with autonomous electric vehicles operating in public areas.

In this research, a telecommunications system for autonomous electric vehicles is built by utilizing the existing telecommunications infrastructure, which is WiFi 5 GHz (802.11ac). This infrastructure selection is based on the availability of 5 GHz WiFi, commonly used as a WLAN device. In addition, the use of 5 GHz WiFi is considered close to the standard frequency for vehicle communication systems (802.11p) with devices that are easily found in the development area. This telecommunication system supports the functionality of autonomous vehicles to communicate with infrastructure and the surrounding environment, particularly for moving, adjusting position and direction, and avoiding collisions, which is commonly known as the vehicle-to-infrastructure (V2I) communication system.

B. Object Detection and Recognition System for Autonomous Electric Vehicle

The development of object detection and recognition systems for autonomous electric vehicles is currently being carried out by several related industries worldwide. This is an effort to support the birth of private or mass vehicles that are environmentally friendly, operated without too much demand for the control role of humans or their passengers. Not only abroad, but the presence of autonomous electric vehicles in Indonesia is also highly expected, especially in response to the government's plan to introduce electric vehicles and their autonomous versions to be operated in the new capital city within the next four years. The development of autonomous electric vehicles is also supporting the national research agenda for electric cars in the future, continuing the national electric car research carried out several years earlier. The autonomous electric vehicle developed by LIPI is slightly different from similar vehicles abroad.

The autonomous electric vehicle will be intended for a limited environment with a small passenger capacity such as a golf cart to a large passenger capacity such as a bus. Basically, the vehicle needs a reliable and appropriate object detection and recognition system to run properly and safe, regardless of the autonomous electric vehicle type that will be developed. Therefore, in this research, object detection and recognition system are designed and developed to support the mobility of autonomous electric vehicles. The system is built by utilizing the information processing methods sourced from omnidirectional images, point cloud data LiDAR (Light Detection and Ranging) and thermal images. And also employing methods of detecting and recognizing objects based on the concept of deep learning, which is one of the variants of artificial intelligence technology (AI). The concept of AI-based technology is also a supporting element for the industrial revolution 4.0. Its implementation is strongly encouraged by the Ministry of Industry through the "Making Indonesia 4.0" initiative. One of the agendas is to make Indonesia a major producer of conventional and electric motor vehicles in southeast Asia.

Sivitas Terkait : Nasrullah Armi,