×
Community Blog Alibaba Cloud's "Big Computing Power" Speeds Up Inceptio's Autonomous Driving Simulation by 20!

Alibaba Cloud's "Big Computing Power" Speeds Up Inceptio's Autonomous Driving Simulation by 20!

This short article discusses Inceptio Technology's usage of data intelligence systems through its relationship with Alibaba Cloud.

On May 24, Inceptio Technology said in an interview that the company has carried out in-depth cooperation with Alibaba Cloud in the field of data intelligence systems. Based on cloud computing, Inceptio has built an industry-leading high-concurrency and high-elasticity infrastructure platform of data storage, computing, and scheduling to accelerate autonomous driving simulation and reduce computing resource consumption.

According to statistics, China's current market size of trunk logistics is about 4 trillion yuan, showcasing great potential. However, the low operating efficiency and high cost of large truck teams are the pain points of the industry. Inceptio Technology (established in 2018) has solved the problems of trunk logistics mentioned above through autonomous driving technology and operation.

1

The R&D, testing, and operations of autonomous driving generate large amounts of data. According to rough calculations, if a road test car carries out full collection, it will generate more than 2 TB of data in one day. As mass-produced vehicles continue to be delivered at scale, the amount of data will grow exponentially. Using data from real scenarios of trunk transportation can reduce R&D investment in autonomous driving and improve R&D efficiency. Exploring values from massive data in massive data requires big computing power.

The article takes the deep neural network model required by the Xuanyuan autonomous driving system of Inceptio Technology as an example. The development of the model relies on an advanced, elastic, and stable AI training environment on the cloud. Necessary perception models and decision algorithms required by autonomous driving also need large amounts of data for continuous iteration and updates. Alibaba Cloud provides a training environment on the cloud for Inceptio but also opens hundreds of cloud servers with the most advanced GPU accelerators to speed up AI computing and shorten the iteration cycle of products.

According to reports, Inceptio Technology and the Alibaba Cloud Container Service for Kubernetes (ACK) Team are exploring cloud-native AI technologies to speed up the R&D of AI algorithms. The development efficiency of AI algorithms is improved by deploying a flexibly customized and on-demand containerized Jupyter notebook development environment.

Under the compute-storage separation architecture, Inceptio uses Fluid cloud-native data orchestration and acceleration to optimize data access performance and improve resource usage. The ACK Serverless ECI elasticity and cloud-native data lake solutions satisfy the high-concurrency and low-latency large-scale elastic resource supply of simulation experiments of Inceptio. This accelerates autonomous driving simulation by more than 20 times and saves about 30% in computing resource consumption.

In the future, Inceptio and Alibaba Cloud will deepen business cooperation and help Inceptio build a unified service platform of capacity operations and management to serve the trunk transportation logistics system. Through the full-stack L3 and L4 autonomous driving technology, Inceptio Technology provides logistics customers with a new generation of safer, more efficient, and lower-cost Transportation-as-a-Service (TaaS) freight networks. As of the end of April 2022, the commercial operating mileage of autonomous driving of smart heavy trucks equipped with the Xuanyuan system of Inceptio exceeded 2 million kilometers.

0 0 0
Share on

Alibaba Cloud Native

93 posts | 6 followers

You may also like

Comments

Alibaba Cloud Native

93 posts | 6 followers

Related Products