QUT seeks industry-government alliance to enable self-driving cars

Researchers at the Queensland Institute of Technology (QUT) are promoting the use of artificial intelligence to determine the feasibility of self-driving cars on Australian roads.

The QUT Robotics Center has conducted a research project on the mapping of self-driving cars using AI. Professor Michael Milford, Deputy Director of the Center, said updating the map is a major challenge for the adoption of self-driving cars.

Milford said Australia has an opportunity to catch up quickly given that mapping is not a mature area globally.

“Current out-of-the-box European mapping solutions are unaware of Australia’s unique signs and infrastructure and require customization,” he said. “Self-driving cars are a little further away, but the main goal now is to ensure that digital, physical, and regulatory infrastructures are available.

Reference: Robots supporting researchers in the Antarctic Conservation Program

“When we realize that technology doesn’t serve its intended purpose, we need to plan and design the technology for the purpose from the beginning, rather than making it a shoehorn at the end.”

Milford has a vision of map creators, localization services that let vehicles know where they are on the map, and work done in collaboration with the government to update infrastructure and regulate privacy.

Specializing in robot and self-driving car positioning research, the QUT Center is collaborating with government and industry on the future of HD maps by investigating ideal models of government-industry collaboration.

“For example, positioning systems will be difficult to work well unless the car is clearly aware of changes in the environment, such as road construction,” he explained.

“Government notices about these events will potentially be very important and require meaningful involvement or oversight due to the important data and privacy implications of these maps.”

The professor added that there is something to be done to update the positioning system. He said current positioning systems work well in most cases, but have the drawback that technologies like heavy rain and tunnels are still unreliable enough.

“There’s nothing worse than having a car that you think is in one place, but actually in another, and misreferencing the wrong section of the HD map as a result of that positioning error,” Milford said. Mr. says.

“Now, if we start a step-by-step approach towards this collaboration model, within two years, information from private map stakeholders, governments, and even vehicles on the road, all these major We have a working prototype of how to share it among our stakeholders, to ensure that the map is as accurate and up-to-date as possible. “

Related coverage

Will the pandemic map a new course for self-driving cars?

In the future, where human contact requires stronger protection, self-driving cars could give people better control than ever before.

We have been promised self-driving cars for years. So why aren’t they on the road yet?

The technology behind self-driving cars exists, but everything else hasn’t happened yet.

This is how self-driving cars are becoming a smarter surgical blueprint.

The same technique in your cockpit can affect future surgical procedures.

The cars of the future are connected, autonomous, shared and electrical

The automotive industry is undergoing a major digital transformation. By 2030, cars will be more connected, autonomous, shared and electric.

QUT seeks industry-government alliance to enable self-driving cars

Source link QUT seeks industry-government alliance to enable self-driving cars

Back to top button