zorazhuang for Getty Images
“2018 marks the first year of commercialization for autonomous driving,” Baidu’s Chairman and CEO Robin Li
The automobile is going autonomous and this shift has made way for technology players along with the traditional OEMs (original equipment manufacturers). Here, we'll take a look at Baidu in particular, and its collaborations, efforts, and achievements.
Baidu (BIDU) began its research and development on autonomous driving back in 2013 with its Institute of Deep Learning leading the way. By 2015, it was performing a few successful road tests. In 2016, they created the Autonomous Driving Unit (ADU) in Silicon Valley to address the technical and business challenges.
Baidu reached an important milestone in its journey towards driverless vehicles with the launch of Apollo in April 2017. Apollo is an open and secure software platform to propel autonomous vehicle technology and ready for adoption by its partners. It is a complete service solution that includes a vehicle platform, hardware platform, software platform, and cloud data services.
By September, Baidu released an upgrade built on Apollo 1.0, dubbed as Apollo 1.5, with five additional core capabilities including obstacle perception, planning, cloud simulation, High-Definition (HD) maps, and End-to-End deep learning. Additionally, a $1.5 billion (10 billion RMB) Apollo Fund was announced for investing in autonomous driving projects over the next three years.
In April 2018, Baidu unveiled the upgraded version of its open autonomous driving platform called the Apollo 2.5—which is equipped for a geo-fenced, high-speed autonomous driving scenario.
The adoption of autonomous vehicles requires sound automotive cybersecurity. To work in this regard, Baidu is setting up a cybersecurity lab in collaboration with China Automotive Technology and Research Center and China Academy of Information and Communications Technology Beijing’s Haidian district.
Continuing with its fast pace of developments, Baidu announced Apollo 3.0 In July 2018. Apollo 3.0 offers new solutions that support valet parking, autonomous mini buses and microcars.
“Apollo 3.0 marks a new era for the volume production of autonomous vehicles,” said Zhenyu Li, Vice President and General Manager of Baidu’s Intelligent Driving Group.
Click to open in new window
Source: Apollo website
Baidu’s initiatives aren’t restricted to cars. Building upon its strategic agreement with King Long United Automotive Industry Co., Ltd. in October 2017, Baidu has partnered with Softbank’s subsidiary—SB Drive for developing and deploying "Apolong" — self-driving mini buses for the Japanese market. Apolong is an Apollo-powered, level 4 mini bus co-developed by Baidu and King Long. The first fleet of these buses will be deployed in geo-fenced areas across five Chinese cities.
The first lot of buses will be exported to Japan in beginning of 2019. This will just be the beginning of a great opportunity given that Japanese Prime Minister Shinzo Abe has his eyes on self-driving vehicles as revealed by his remarks, “Do please come and see the 2020 Tokyo Olympic and Paralympic Games. It will be like Buy one, get two because I can tell you that in 2020 Tokyo, self-driving cars will be running around, and you will be able to use them to move about.”
These driverless vehicles can help resolve issues such as driver shortage faced by Japan's public transport.
While great progress has been made by Baidu, challenges remain. To deliver safer self-driving technology, Baidu and Intel (INTC) recently announced the integration of Mobileye’s Responsibility Safety (RSS) model in both the open source Project Apollo and commercial Apollo Pilot programs. Additionally, Mobileye’s surround computer vision kit will be employed for visual perception solution.
The tie-up between Baidu and Intel attempts to strengthen the safety aspect of autonomous vehicles. Mobileye’s RSS provides safety assurance of AV decision-making by “formalizing common-sense human-centered concepts of what it means to drive safely” like maintaining safe distance while following a vehicle or the concept that right-of-way is given, not taken.
Intel estimates the vehicle systems, data and services market opportunity to be up to $70 billion by 2030. Artificial Intelligence (AI) is at the center of autonomous vehicle technology which operates ‘probabilistically’ and is thus prone to mistakes when it comes to decision making. To combat such shortcomings, Mobileye recommends “adding a separate, deterministic layer on top of AI-based decision-making solutions in autonomous vehicles.” Together, Baidu and Intel will accelerate work in this segment.
China is a booming market for autonomous vehicles with an aim to have at least 50% of all new vehicles to be ‘partially’ or ‘fully’ autonomous by 2020. Reduction in traffic accidents, lowering fuel consumption and emissions are some of the motives that the government is looking to fulfill with its encouragement to autonomous driving vehicle technology. China makes up about 30% of the global passenger vehicle market and is expected to become the world's largest market for autonomous cars.
Baidu is leading China’s push towards driverless technology and can boast of over 116 partners from across the globe, including names like Microsoft, BAIC, Pioneer, TomTom, ZTE, NVIDIA, Daimler, Valeo, Bosch, Ford, BMW and Intel. With its growing network of partners, focus on safety and development of core technology, Baidu is consolidating its position in the software-defined automobile world.