Baidu has released second generation of its open-source autonomous driving platform Apollo 1.5. Building on Apollo 1.0, Apollo 1.5 opens up five additional core capabilities which include obstacle perception, planning, cloud simulation, High-Definition (HD) maps and End-to-End deep learning, providing more comprehensive solutions to developers and ecosystem partners to accelerate the deployment of autonomous driving.
The obstacle perception capability enables vehicles to accurately identify obstacles during both day and night. With the planning capability, vehicles can plan the correct driving path and make optimal driving decisions. The cloud simulation system on Apollo is the only platform that provides open, built-in HD maps tailored for autonomous driving. Based on large scale cloud computing capacity, Apollo has a vast amount of real traffic data in China, and has a virtual running capacity of one million kilometers per day.
Apollo, announced in April, is an open platform that provides a comprehensive, secure, and reliable all-in-one solution that supports all major features and functions of an autonomous vehicle. It has attracted 70 global and Chinese partners, including OEMs, Tier 1 suppliers, developer platforms and technology start-ups.
Over the past 2 months, Apollo has seen dozens of code updates each week and the addition of more than 65,000 lines of new code. To date, more than 1,300 companies have downloaded Apollo source code and nearly 100 companies have applied for open data via the Apollo website.
Baidu’s release of Apollo 1.5 was also occasioned by the announcement of a 10 billion RMB ($1.5 billion) fund by the company. The ‘Apollo Fund’ will invest in 100 autonomous driving projects in the next three years.