In Haoxin Wang’s eyes, The once futuristic vision of self-driving cars is closer to reality than many think. His research is focused on revolutionizing the design of these self-driving cars.
Recent advances in autonomous driving technology range from adaptive cruise control and braking to the development of fully autonomous vehicles.
An assistant professor in the computer science department at Georgia State University, Wang addresses the need for advanced vehicle computing power to make self-driving technology more accessible and efficient.
According to Wang, the key lies in edge computing and sustainable artificial intelligence (AI), which could improve the behavior of smart vehicles.
“We want to make sure all vehicles have enough computing power when running AI applications such as autonomous driving algorithms and image processing,” said Wang, adding that the work is called “protocol and algorithm design”.
“Our vision is that in the future, all vehicles will have fully automated driving capabilities.”
Haoxin Wang, an associate professor of computer science, envisions a future where all vehicles have fully self-driving capabilities.
This means a more equitable and equitable computing environment, allowing vehicles to be equipped with cutting-edge technology regardless of price.
Through research, Wang and Jiang Xie at the University of North Carolina at Charlotte can transform the capabilities of intelligent vehicles through automotive edge computing.
Edge computing refers to systems of servers and wireless technology that act like remote computers in vehicles. This approach allows computational work to be offloaded from the car to external resources.
Recent advances in automotive technology rely on either downlink connections (transmission of information from a computer server to the vehicle) or uplink connections (transmission of information from the vehicle to the server). Downlink connections support features such as streaming entertainment. On the other hand, functions that require the vehicle to capture information from the environment, such as autonomous driving technology, are supported by uplink connectivity.
In some cases, both downlink and uplink connections are used. For example, navigation technology uses uplink technology to send location data from your car to a remote server, and downlink technology to send updated route information back to your car.
However, current vehicles have limited computing power. As a result, manufacturers must choose how to allocate resources to support each of these functions.
Traditional network resources are considered asymmetric. In other words, allocate more resources to downlink applications than to uplink applications. In the future, this could become an issue as uplink applications are integral to autonomous driving technology.
This is where Wang sees opportunities for improvement.
His research proposes a new model of resource allocation involving external edge servers using algorithms that improve the performance of existing edge computing technologies.
“This vehicle will become something of a data collector,” says Wang. “It collects data from the surrounding environment and sends the data to an edge server for processing. Once processed, the edge server returns the results to the car.”
Offloading the computational process to an external server relaxes the limitations of the vehicle’s hardware. This technology could potentially provide greater parity between vehicles, regardless of their built-in computing power.
In addition to making these forms of technology more accessible, Wang wants to make them better for the environment. He speaks in terms of sustainable artificial intelligence ecosystems.
“Most researchers today are most interested in AI performance, or what AI can offer,” said Wang. “What we are interested in now is the sustainability of AI.”
AI applications are data and power intensive, resulting in significant energy consumption and carbon footprint. Wang says he strives to make AI more environmentally friendly, so perhaps with the help of his research, AI could become more sustainable as it becomes more prevalent in the future.
“Such eco-friendly AI is very important for the future of AI. We need to make it green and sustainable to support our community,” said Wang.
