Deploying scalable vehicle apps to analyze insights from sensor data

This architecture demonstrates how you can deploy scalable apps to a vehicle and then analyze and visualize insights from the data that comes from the sensors in and around the vehicle.

Architecture diagram

High-Level Diagram. — Figure 1. Automotive High Level Diagram

The IBM Automotive architecture has three layers: the vehicle layer, the connected vehicle platform layer, and the enterprise hybrid multicloud layer.

The heart of the IBM connected vehicle AI platform is the mobility services application, which propels the vehicle by using a Human-Machine Interface (HMI). HMI is a haptics- and voice-enabled UX device that fronts the onboard intelligence. People in the vehicle use the HMI to communicate with the vehicle. A version of the HMI is available as a mobile app that automatically synchronizes with the onboard capability. Secure workloads are deployed as microservices.

The Automotive architecture has these functional requirements:

Securely communicate with the vehicle from within or outside the vehicle
Offer real-time inference and AI analytics onboard the vehicle for improved in-vehicle productivity
Supplement the Connected Vehicle Platform with the ability to manage services at scale
Analyze data at the source for collision avoidance systems and predictive maintenance

The challenges in the realm of mobility services and vehicle-to-everything (V2X) communication are unique:

Support for high throughput and real-time telematics data ingestion, process, and analytics
- Through analytics of the telematics data, promote driver safety and navigate through better driving conditions on the road
- Help to increase lifestyle efficiency
- Help to avoid collisions, reduce insurance costs
Provide comprehensive and flexible APIs to enable application developers to implement cross-domain business applications.

A connected vehicle depends on its supporting networks. Many networks are involved in the architecture, both inside and outside the vehicle. The networks that are inside the vehicle have unique protocols and a wide range of data rates. On the low end of data rates, the Local Interconnect Network (LIN) is used for low-speed applications like sensors and actuators at 20 kbps. On the higher end of data rates, Ethernet is used for high-speed applications such as infotainment and advanced driver-assistance systems (ADAS). Wireless interfaces, such as 3G, 4G, future 5G, BT, wifi, and V2X, operate at 100 Mbps to gigabit speeds.

Nonfunctional requirements

Automotive architectures must satisfy several nonfunctional requirements.

Security: Every connection in and out of the connected vehicle must use a secure protocol. All data in transit must be sent over secure protocols and any data that is stored in the cloud or data center must be encrypted. When connected vehicles and related devices are onboarded, use keys, certificates, or tokens.
Response time: Speed is synonymous with mobility. MaaS architectures require minimal latency because decisions must be made quickly whether the communication is between two vehicles or between a moving vehicle and infrastructure.
Connectivity: MaaS is a special case of connected cars, and connected cars require on-demand connectivity. The connectivity can be local, to the cloud, or to a satellite.

Availability, maintainability, and scalability must also be addressed in a MaaS architecture.

Components

The following represents a typical set of components that are deployed in a connected vehicle solution.

Table 1. Components
Architecture components	How the component is used
Sensor	The sensor, which provides a measurement that is meaningful for the physical entity (for example, location, temperature, humidity, traffic, and vehicle sensors). It might or might not be physically attached to the physical entity. The sensors and devices include built-in instrumentation such as RFID trackers, weight tracking, intelligent LED lighting, and others that acquire data or are controlled intelligently.
Connected car	Physical vehicle is connected with the sensors and the internet.
Driver or Passenger	Person sitting in the vehicle.
Real-time decision	Acquire data from sensors and systems for analysis and real-time decision making.
HMI	A Human-Machine Interface (HMI) is a user interface or dashboard that connects a person to a machine, system, or device. It provides entertainment to the passengers and more control, safety, and precision to the drivers.
In-vehicle API	Common set of services that are needed by Mobility as a Service, for example, e-commerce portals for pizza delivery, communication, Bluetooth, vehicle navigation, and onboard diagnostics.
Security services	Services that provide access management to the vehicle.
AI services	Services that allow data scientists, developers, and retail business domain experts to drive improved business results and customer experience through the infusion of AI. With the AI services you can securely use your own data and to use licensed and public external data in a governed way to create, train, and deploy self-learning models.
Gateway	An IoT gateway that is typically run on premises (in a plant or warehouse). The gateway is responsible for connecting SCADA systems, sensor gateways, and devices to the IoT platform. To do this, it receives device data from device and industrial protocols, formats the data as needed, and uses an IT protocol to communicate with the platform. It might also do filtering and aggregation of the data.
Network services	Provides network capability to deliver content through the Internet (DNS, CDN, firewall, and load balancer).
Edge services	DNS, firewall, and load balancer.
Data and AI platform	Contains machine learning (ML) and AI workloads, data analytics, model lifecycle management, visual services, BigData, data warehouse, and data quality.
IoT transformation and connectivity	Data IoT transformation and connectivity services, API connectivity, and data connectivity.
HMI mobile application	Human Machine Interface application. Mobile, iPad, or any device that can run applications.
Vehicle services	V2X, Software Services (OTA) such as upgrade and maintenance, diagnostics, battery health, and vehicle health.
Live conversation	Live conversation and chatbots for car manuals, feature functions, car assistance, and help. Connected to Peer Services to order the services on a third-party cloud.
Peer services	IBM or third-party cloud system that provides services to bring data and capabilities to the social platform. Connect and order any services directly from the car through peer services, such as McDonalds and Starbucks, subscription services like Pandora, and third-party services like Netflix and Spotify.
Connected vehicle insight	Services such as context map, driver behavior, data transformation, real-time analytics, geo-spatial analytics, and BigData.
Data services	Brings all the third-party external data on the platform, like Twitter and weather data.
Mobility services	Application runtime. Supports application logic, which is part of an AI Solution that is built on the Connected Vehicle platform. The platform interacts with information in the data lake, external services, and workflow as needed. Examples are mWallet, concierge, infotainment, mServices, car sharing, and taxi.
User	User who is connected to a web application or by using a mobile application.
Transformation connectivity	Classical service bus functionality for decoupled integration of systems and machines, taking over integration logic, systems, and machines. Provides relief from integration-specific logic and enables standardization of maintenance and release management.
Connected Car (CC) Registry	A database of registered vehicles that is available to authorized entities.
Data stores	Provides the persistence layers to power the cloud app to store, collaborate, visualize, share, and gain insights from data.
Management hub	Recognition, patch, lifecycle, compliance, and security management of edge devices, industrial robots, and automobiles.
Third-party applications	Third-party applications such as security apps or reporting apps that might be integrated as part of the solution.
Enterprise applications	Applications that accomplish business goals and that can interact with cloud services.
Edge governance	Provides and enforces the appropriate in-service lifespan of devices. Plans a smooth, nondisruptive, and secure changeover as new systems and capabilities are introduced.
Security	Security is pervasive through all layers of the architecture. Existing systems (such as MES, SCADA, and devices) already have their own security models. Manages both IT and Operations Technology security elements.
Multicloud Management	A multicloud management solution.