A battery is a key component in an electric vehicle because it enormously influences user experience. OEMs try to monitor and analyze a product’s battery condition for finding insights while an end-user is using it. As battery life is variable by usage environment and habit, OEMs reflect those insights into battery life enhancements. This article is an example of an end-to-end solution for battery lifecycle management in mobility industry. It discusses how to manage each user’s battery lifecycle through Azure Digital Twins.
- A product/device gathers and sends telemetry data such as battery’s current, voltage, SOC (State of Charge), location, etc. to IoT Hub. In terms of device management and operation, OEMs could use their own telematics platform or develop a consumer telematics solution based on Azure IoT SDK, Azure Sphere, Azure/Free RTOS and Azure IoT Edge.
- Azure Functions distributes messages to each product’s Azure Digital Twin instances. A message needs to include identification such as product ID for routing and updating a message to a right twin instance through Azure Digital Twins APIs.
- Azure Digital Twins receives messages in Azure Digital Twins models, ontology and handles event notifications using Azure Event Grid.
- Azure Functions and Azure SignalR services push digital twin telemetry data to connected clients such as a single web page, a mobile application and Azure Maps. Azure SignalR supports updating real-time metrics and status to clients without polling servers or submitting HTTP requests.
- Azure Digital Twins provides an integration feature, data history, that allows Azure Data Explorer historize property updates and time series data.
- Once ingested Azure Digital Twins data into Azure Data Explorer, a data analyst or scientist can query historical data of Azure Data Explorer time series database using KQL (Kusto Query Language) with a chart or graph. Azure Digital Twin query plugin for Azure Data Explorer lets users combine and query data across the Azure Digital Twins graph and Azure Data Explorer time series databases.
- A dashboard with near-real time updates gives a user an insight that helps decide on industry KPIs such as SOC (State of Charge), SOH (State of Health) and RUL (Remaining Useful Life) of a battery. Power BI, Grafana and Azure Data Explorer Web UI allow to connect various data sources and visualize analytics dashboards.
- Data scientists can build machine learning models that predict battery conditions such as SOH and RUL using advanced analytics tools, Azure Machine Learning and Azure Data Bricks.
- The machine learning models can be deployed to Azure Services, Azure Container Instance, Azure Kubernetes Service and Azure Functions. Azure Digital Twins sends near-real time telemetry to a machine learning model and gets back an inference result and update it in Azure Digital Twins’ property through Azure Functions.
- Azure Sphere: Build solutions that start in silicon and securely connect, manage, and protect new and existing devices.
- Azure RTOS: Operating system for an embedded development and providing reliable, ultra-fast performance for resource-constrained devices.
- Azure IoT middleware for FreeRTOS: Provide a proper SDK to a developer for connecting devices to Azure.
- Azure IoT Edge: Edge runtime enables custom logic on edge device and provides communication operations to Azure
- Azure IoT Hub: Cloud IoT gateway that connects, monitor and manage a large volume of message and IoT assets
- Azure Functions: Event driven serverless compute platform allows end-to-end development with multiple programming languages
- Azure Digital Twins: A digital representation of a battery that provides a near-real time status and standards data to downstream services
- Azure Event Grid: Handling Azure Digital Twins’ events at massive scale
- Azure SignalR Service: A broker service enables real-time communication into a web application
- Azure Maps: A suite of geospatial service visualizes product’s location and information in an application
- Azure Web Apps: Build, deploy and manage a web application with reliable scalability
- Azure Data Explorer: A fully managed data analytics service enables real-time analysis on telemetry and historical data of Azure Digital Twins for getting trends and an insight
- Azure Data Lake: Store big data and make it easy to use for data scientists, analysts and developers
- Azure SQL: A fully managed relational database service focuses on application development
- Power BI: Business intelligence platform connects and visualize various data to build a dashboard for getting data insight
- Azure Managed Grafana - Grafana dashboards that is provided as a fully managed Azure service
- Azure Data Explorer Web UI - a fast and highly scalable data exploration service for log and telemetry data
- Azure Machine Learning - an enterprise-grade service for the end-to-end machine learning lifecycle
- Azure Databricks - Design AI with Apache Spark™-based analytics
- Azure Kubernetes Service (AKS) - Deploy and scale containers on managed Kubernetes
- If you have your own IoT platform or streaming data platform such as Kafka, it can connect to Azure Digital Twins through Azure Functions.
- Instead of Azure SignalR, MQTT broker service enables real-time communication to web applications whenever Azure Digital Twins is changed.
- There are professional solutions for battery life prediction. If you simulate battery systems with 3rd party solutions, you can integrate it with Azure Digital Twins for Advance Analytics.
- Integrate Azure Digital Twins with 3rd party simulation
Modeling a battery ontology of Azure Digital Twins needs domain knowledge in a battery and product. It doesn’t need to exactly follow physical product structure. A manufacturer can define their own battery ontology considering of component’s properties and telemetries.
In the case of a vehicle, a battery system consists of pack, modules and cells. The minimum unit is a cell which is container to store energy. Multiple cells are grouped together to form a module. Multiple modules are grouped together to form a pack.
A Novel Prediction Process of the Remaining Useful Life of Electric Vehicle Battery Using Real-World Data, 2021 Dec
BMS (Battery Management System) is a key component that manages the parameters of a battery system. It communicates and manages cells, modules and others through CAN (Controller Area Network) network protocol. Automotive OEM possesses own know-how on a battery system. It is difficult to standardize it as a common ontology. It needs to be defined through discussion with OEM’s battery expert. OEM can manage each customer’s battery lifecycle with Azure Digital Twins ontology.
Example: A battery system structure
Example: Azure Digital Twins Battery Ontology
- EV battery performance and optimization e.g.) Hyundai Motor Group Pilots Digital Twin with Microsoft
- Connected vehicle management for design and service
- Vehicle sharing and rental service management
- Battery test in engineering
These considerations implement the pillars of the Azure Well-Architected Framework, which is a set of guiding tenets that can be used to improve the quality of a workload. For more information, see Microsoft Azure Well-Architected Framework.
Vehicles move across multi regions. It needs to connect vehicles to nearest cloud region with the lowest latency.
- If you use Azure IoT Hub as cloud IoT gateway, you can use Azure IoT Hub Device Provisioning Service (DPS) for zero-touch provisioning and automatic connection to a right IoT Hub based on your rule definition.
Security provides assurances against deliberate attacks and the abuse of your valuable data and systems. For more information, see Overview of the security pillar. In the case of connecting and tracking end-users’ vehicles, device data is being exposed to public network. Microsoft provides Zero Trust Cybersecurity for the Internet of Things that provides the principals of zero trust security to help proactively establish an IoT strategy. Azure Digital Twins is the central service that integrates legacy systems such as ERP and provides downstream data to other services. For security considerations, see Secure Azure Digital Twins.
Vehicles being used by end-users generate huge amounts of data. Review Cost optimization in your IoT workload for implementing cost effective IoT solutions. In terms of end-to-end solutions, see Cost optimization in the Microsoft Azure Well-Architected Framework.
A large volume of Azure Digital Twins ontology and twin graph could cause high complexity and performance issues based on a requirement. Whenever a system or code queries data in Azure Digital Twins, it calls APIs. For optimizing performance of twin graph traversals, see How to speed up Azure Digital Twins Queries with Caching Strategy.