Huawei’s Auto Ambitions
Huawei’s Rotating Chairman Zhijun Xu spoke at a conference that themed intelligent connected vehicles in Beijing on Oct. 22 and revealed for the first time that Huawei is going to develop LiDAR and millimeter-wave radar in-house. This became the headline of multiple media in the following days. However, after a closer look into Huawei’s strategy, it is more than supplying key sensors on autonomous vehicles. In this article, I go through why Huawei targets the auto industry, the C/C system architecture that Huawei initiated, and the 5 parts of Huawei’s auto strategy based on the C/C architecture.
Huawei’s plan is to do everything short of building a vehicle.
The automotive industry is transforming towards automation, connectivity, electrification, and shared mobility, commonly referred by the acronym ACES. In Huawei’s words it is “the convergence of automotive and ICT industries”. It seems like a natural extension of Huawei’s core capabilities to develop new technologies to enable auto OEMs to build next generation vehicles, in specific intelligent connected electric vehicles.
Also, with over 30 years of experience in ICT, Huawei has probably achieved saturation on many business lines and is ready to take on a new growth engine. Instead of entering the auto manufacturing business like Tesla, which is already crowded and could only reward marginalized profits, Huawei is carving out another value-added space which falls in line with its existing role as a B2B supplier.
ACES translates into a growing demand for software-enabled features on the vehicle and a corresponding need for increased computing horsepower. Therefore the traditional electrical/electronic (E/E) architecture will no longer be viable to support the growth in content and complexity. Tesla is no doubt the leader of architectural innovation with the launch of Model S back in 2012.
Huawei has initiated the computation/communication (C/C) architecture that is distributed networks and 3 domain controllers. With the C/C approach, Huawei expects easy software iteration, hardware replacement, and sensor fusion extendibility on the vehicle, very similar to Tesla.
The C/C architecture, as showed in the diagram, is distributed gateways with 3 domain controllers that are smart cockpit, vehicle dynamic control and intelligent driving. Huawei is building 3 platforms corresponding to the domain controllers called CDC, VDC, and MDC respectively. Together with their connectivity solutions and cloud service, which support the C/C architecture, Huawei is aiming to be the supplier with a full-stack solution to OEMs.
With the overall structure in mind, we will go into details on the 5 business lines forming Huawei’s auto strategy.
MDC Intelligent Driving Platform
An autonomous vehicle will be like a Mobile Data Center (MDC) on wheels that constantly aggregate and process massive amount of sensor data to do planning and control. To meet the requirements on computing platforms, Huawei’s MDC solution aims to provide the bottom-layer chip (Ascend), platform, in-vehicle operating system, and the upper-layer development framework for running algorithms.
On this solid base, Huawei plans to develop 3 ecosystems with external partners.
- Sensor ecosystem: Huawei would encourage external sensors (such as cameras, millimeter-wave radars, LiDARs, ultrasonic radars and GPS) to connect to MDC. As an option, Huawei is going to develop millimeter-wave radar and LiDAR in-house.
- Application ecosystem: Huawei would welcome partners to develop automated driving related applications and algorisms based on MDC.
- Actuator ecosystem: Huawei is to build a standard API for all the actuators like ESC, EPS, VCU to easily integrate onto MDC.
CDC Smart Cockpit Platform
Within the smartphone ecosystem, Huawei already owns a huge user base as the manufacturer of smart devices. They believe that automotive cockpit presents the next opportunity, which would be an ecosystem of hardware, software, and applications dedicated to smart cockpit. Moreover, the experience of smartphone and smart cockpit can be synchronized.
Huawei plans to build a Cockpit Data Center (CDC) on top of their Harmony operating system and the Kirin chipset. It will be an open platform to support 3rd party hardware like screen, MIC, audio and 3rd party applications similar to the ones on smartphones. Within the smart cockpit ecosystem, hardware can be replaced, new applications can be rolled out and software iterated constantly, all together creating an advanced in-vehicle user experience.
VDC Intelligent Electrification Platform
Huawei plans to develop a MCU on top of which to build a Vehicle Dynamic Control (VDC) system. Automakers can customize based on the VDC to create differentiated user experience.
By leveraging its capabilities in the energy sector, Huawei also plans develop products and solutions around electric power systems, charging and battery management, ultimately solving the performance and cost issues.
Intelligent Connectivity Solutions
Huawei believes their 5G in-vehicle communication modules, T-Box, Ethernet switches together with OceanConnect form a comprehensive solution to enable the ultimate high-speed low-latency connections between vehicles and vehicles/infrastructures/the cloud.
Cloud Service
- Octopus: it is a cloud service dedicated to the training, simulation and test of autonomous driving.
- OceanConnect: it is the Internet of Vehicles (IoV) platform designed to help vehicles communicate with each other and the environment. Huawei claims that it is linked to global public cloud architecture and supports hundreds of millions of connections and millions of high concurrent connections.
(Initially published on LinkedIn on October 30, 2019.)
