Home / SmartTech / Qcraft raises over $ 24 million to train autonomous vehicle systems in simulations

Qcraft raises over $ 24 million to train autonomous vehicle systems in simulations

Qcraft, a robotics startup focused on autonomous vehicle (AV) development, announced this morning that it had raised over $ 24 million in seed capital. (A spokesman would not disclose the exact amount, but said it was in "tens of millions.") The company says it will use the capital to further develop its large-scale intelligent simulation platform that generates driving scenario data that Helping engineers Optimize the motion planning and decision-making functions of autonomous systems.

A simulation-based approach to AV development is not new – Waymo, Uber, Aurora, GM Cruise, and Argo AI are startups that use digital environments to optimize settings. AI underpins their driverless cars. However, Qcraft claims that its simulation system and self-learning framework are more sophisticated than most others, with a focus on use cases in complex urban environments. According to Qcraft, only three steps are required in practice ̵

1; mapping, simulation and validation – in order to master real driving scenarios with his toolkit. As a result, the test costs are "immensely" reduced and development efficiency is increased.

"Unlike the conventional roadmap of autonomous driving companies [we are] to make self-driving cars completely autonomous in a unique way," said Qcraft in a statement. “The holistic solution to this challenge is to find and solve the problems by collecting a large amount of real test data and repeatedly manually adjusting the algorithm parameters. In real-world scenarios, data acquisition would require large fleets and millions of miles of real drive data, and the solution is very expensive and simply not feasible. “


Qcraft's sensor fusion and synchronization supposedly guarantees“ better ”and“ faster ”object detection and at the same time supports the functions of various sensor types such as cameras, lidar sensors, radar devices and inertial measurement units. On the simulation side, the company's software can transform key cases from real data. To this end, Qcraft says that its simulation-trained vehicles can autonomously drive on highways and park in parking lots, and it has recently shown that they can navigate “complex scenes” like fast-food passages.

  VB TRansform 2020: The AI ​​event for managers. San Francisco July 15-16

Qcraft’s autonomous driver system is said to operate on a number of vehicle platforms, as well as business applications such as robo-taxis, grocery vehicles, and Tier 3 automobiles. (The Society of Automotive Engineers defines Level 3 vehicles as those with environmental recognition capabilities that can make some decisions for themselves in geofenced locations, but still require human oversight.) The company is currently testing 10 Level 4 autonomous vehicles (most of them Coping with scenarios) Qcraft was founded in March 2019 by four Waymo engineers: co-founder Wang Kun, who worked for three years. He needs a DMV driver's license to drive it on public roads.

Qcraft was founded in March 2019 with Waymos simulation team as a software developer; Yu Qian, former technical director at Google Maps and Waymo; and Hou Cong and Da Fang, engineers who previously led Waymo's perception and movement planning teams. The rest of the company's workforce – around 130 employees in research and operations centers in Beijing, Shenzhen and Silicon Valley – comes from leading self-driving companies such as Tesla, Uber and Ford, as well as from technology giants such as Nvidia and Facebook.

IDG Capital led the start with Vision + Capital and Tide Capital.

Sign up for Funding Weekly to start your week with VB’s top funding stories.

Source link