GIGABYTE PILOT: The Autonomous-Driving Control Unit at the Heart of Taiwan’s First Self-Driving Bus

Autonomous vehicles have hit the streets. Case in point: WinBus, the self-driving electric shuttle bus developed by Taiwan’s Automotive Research Testing Center (ARTC), a government-funded non-profit research facility. Since August of 2019, the sleek and silver double-ended bus has been making the rounds at the Changhua Coastal Industrial Park in central Taiwan, among other places. The “robot pilot” at the wheel is the GIGABYTE PILOT—an AI-enhanced embedded in-vehicle computing platform made by GIGABYTE Technology.

Learn More:
《Meet GIGABYTE PILOT: GIGABYTE’s Automated-Driving Control Unit》
《Glossary: What is Artificial Intelligence (AI)?》

The WinBus can hold up to 15 passengers and has a range of 70 kilometers on a full charge. Its unique double-headed design helps it navigate narrow roads where it is difficult to make U-turns. There is no steering wheel, gas pedal, or brake inside the bus; this is because WinBus has achieved Level 4 (high automation) on the “Six Levels of Driving Automation” as defined by SAE International, an automotive standardization body. This means the WinBus can operate without human input as long as certain conditions are met—in this case, the criterion is the fixed route that the shuttle bus follows as it services commuters and tourists.

The WinBus is the distillation of the best that Taiwan’s automotive and high tech sectors have to offer. It is a proud example of made-in-Taiwan excellence, and it comprises components from over twenty Taiwanese manufacturers—including onboard cameras, radar, and other sensors; telematics devices; batteries and power supply units; gearboxes and chassis. At the center of all this is the GIGABYTE PILOT, an Autonomous-Driving Control Unit (ADCU) that serves as the AI Mobile Edge Computing Platform that connects everything together.

Glossary:
《What is Telematics?》
《What is Edge Computing?》

Autonomous vehicles use AI to interpret sensory data and navigate their surroundings. The AI algorithms are often conceived in data centers and fine-tuned through test drives that utilize computer vision and deep learning to continuously improve the AI. GIGABYTE’s server solutions are used by world-renowned developers of autonomous driving technology during the AI training process.

Learn More:
《GIGABYTE Helps Israeli Company Construct the Brain of a Self-Driving Car》
《Glossary: What is Data Center?》
《Glossary: What is Computer Vision?》
《Glossary: What is Deep Learning?》

A different sort of computer is needed to deploy the AI. The vehicle’s onboard computing platform must be powerful enough to process information from the sensors, run the AI, and provide instructions to the vehicle’s control system. Data collected from each drive can also be used to retrain and improve the AI. All this requires a lot of processing power.

GIGABYTE’s ADCU is the optimal option for these computation-intensive tasks. It pairs the customer’s choice of Intel® Xeon® or 9th Generation (or higher) Intel® Core™ CPUs with NVIDIA GPUs or Intel® Movidius™ VPUs to deliver accelerated performance through a process known as heterogeneous computing. The high number of cores allows the ADCU to multitask at a level beyond what human drivers can hope to achieve. The ADCU is also compatible with multiple operating systems (OS). The GIGABYTE PILOT is the ideal platform for the deployment of self-driving AI algorithms.

Glossary:
《What is GPU?》
《What is Heterogeneous Computing?》
《What is Core?》

Once you have a sophisticated AI system, you need input from sensors to help the AI make the correct decisions; you also need the control system to execute those decisions. For this reason, the ADCU needs to be hooked up to onboard sensors such as cameras, lidar, and radars. In addition, it needs to connect to the vehicle’s control system, so it can convert throttle input to torque requests, operate the actuators, monitor the safety system, and so on. In essence, the ADCU is the hub between the vehicle’s sensors and control system.

The GIGABYTE PILOT provides high-speed interfaces to connect with sensors and a full array of I/O ports to support peripheral devices. It utilizes 4G LTE and 5G to allow for real-time system monitoring and the transmission of data to control centers on the cloud for big data analysis—therefore achieving the goals of a modern V2X network. It is compatible with a variety of global navigation satellite systems (GNSS), such as GPS, Galileo, GLONASS, BeiDou Navigation Satellite System, among others. These features ensure that the autonomous vehicle has all the data it needs to help its passengers travel in comfort and safety.

Glossary:
《What is 5G?》
《What is Cloud Computing?》

The GIGABYTE PILOT is designed for reliable operations in harsh environments. It adheres to automotive industry standards and can operate at wide temperature ranges. This is due in part to the ADCU’s highly efficient thermal exchange design, which allows the computer to maintain peak performance without the need to use cooling fans.

To cope with the jolts and vibrations of a vehicle in motion, key components inside the ADCU employ a special rugged design. This is important because, as advanced driver assistance systems (ADAS) improve and driverless cars inch closer to full autonomy, onboard computers must be able to withstand any environment that the vehicle may need to operate in.《Glossary: What is ADAS?》

It is also worth noting that the GIGABYTE PILOT features a built-in smart ignition power control system that supports a wide range of operating voltage, from 9Vdc to 36Vdc. This is necessary to maintain the operational stability of the electric vehicles. Since its launch, the WinBus has traveled thousands of kilometers and serviced thousands of passengers, and one reason it has performed so wonderfully is due to the GIGABYTE ADCU that is at the heart of the autonomous vehicle.

The success of WinBus proves the GIGABYTE PILOT is not only ideal for self-driving cars, but also other types of vehicles or vessels that must operate without human input. Based on its extensive experience in product engineering and manufacturing, GIGABYTE provides end-to-end service to help its partners in various sectors develop ideal autonomous applications from the drawing board to mass production.

For example, autonomous vessels and yachts, automated heavy machinery, automated mining and farming equipment, and diversified autonomous mobile robots (AMRs) used for industrial automation, smart logistics, construction, and the Industrial Internet of Things (IIoT)—all of these can be controlled and navigated with GIGABYTE PILOT.

GIGABYTE Technology is glad to have contributed to the development of WinBus, Taiwan’s first self-driving electric shuttle bus to reach the fourth level of driving automation. As server products and smart city solutions infuse more and more aspects of our lives with the power of computing, we can all expect to lead more comfortable, convenient, and productive lives. This is in keeping with the GIGABYTE motto: “Upgrade Your Life”, which is a vision of the future where high tech will change the way we live, work, and play—for the better.

Learn More:
《GIGABYTE’s ARM Server Boosts Development of Smart Traffic Solution by 200%》
《0 to 100 KPH in 3.3 Seconds! NTHU Builds Electric Formula Student Race Cars with GIGABYTE》