Build an smart cockpit with human, road, vehicle and cloud collaboration
In the context of software-defined cars, an smart cockpit brings a personalized interaction experience to users by integrating "one core, multiple screens, multiple systems", voice recognition, gesture control, LCD instrumentation, HUD, DMS and ADAS integration, etc. It is an important differentiation factor for car manufacturers in the fierce market competition, and also an important differentiation point for car manufacturers to subsequently obtain. It is also an important distinguishing selling point for car companies in the competitive market, as well as a way for car manufacturers to obtain user data and monetize traffic data.
Challenge
Smart cockpit product software and hardware structure are complex
Under the smart cabin scenario, OEMs and suppliers will have a more diversified and open relationship, and suppliers will need to develop compatible hardware and software according to the customer’s needs. At the same time, the smart cockpit software also involves the operating system, middleware, UI design, and other aspects.
Operating system, virtualization, software development capabilities are not enough
Host manufacturers have insufficient software development capabilities for SoC, Hypervisor virtualization, in-vehicle OS, middleware, and upper-layer application software.
Lack of end-cloud convergence solution capability
Lack of end-to-end convergence solution capability for edge-vehicle side, cloud side, and cloud-side collaboration need to carry a set of mature cloud-side collaboration architecture and components.
Solution
EMQ edge components can be adapted to mainstream domain control systems to achieve the edge computing capability of the vehicle domain control side with less resource consumption. The edge component supports multiple data sources access, with edge computing power, time window, DBC interface parsing capability, and multiple data interface output capability.
In the EMQ smart cockpit scenario solution, the EMQ edge computing component is integrated with the vehicle domain control system, using raw data from CAN bus and Ethernet bus as the data source. Combined with the edge stream computing capability of DBC, it can provide dynamically scalable edge engine computing capability through a flexible rule engine to realize command counter-control, write to edge database or log file and cloud-side data forwarding.
After the cloud gets real-time data input from the edge, it uses the powerful central computing power to train large, high-order AI algorithms and inference models. EMQX ECP provides batch update function based on vehicle models, projects, and other dimensions, which can issue trained AI algorithms, decision models and customized data processing rules to the in-vehicle side through a lightweight and secure cloud-edge channel, building an intelligent cabin scenario brain for end-to-cloud collaboration.
Results
- Based on EMQ's cloud-side end-to-end collaborative software architecture, we help car manufacturers build the core capabilities of the smart cockpit with vehicle-cloud collaboration. It realizes software and hardware decoupling capability, edge computing engine, and AI extension capability for cockpit domain control, and builds end-cloud convergence based on big data scenario brain.
- Based on the above capabilities, car enterprises can lower development costs and shorten development cycles to achieve keyless entry, intelligent air conditioning, intelligent seats, scene engines, and other smart cockpit scenarios. This as well as intelligent driving modes, ROTA, chassis tuning, and other personalized customization capabilities can provide a good intelligent driving experience for car owners.