Automotive and Manufacturing Industries - Incident Detection and Response Alarming Best Practices

Automotive Industry

---

Introduction

There are many different workload types within the Automotive Industry. “Connected mobility” is quite a popular one. This workload refers to the integration of technology into transportation systems to improve the flow of traffic and enhance the overall mobility experience. This includes the use of connected vehicles, smart infrastructure, and advanced data analytics to help improve traffic flow, improve safety, and reduce emissions. Connected mobility also includes the integration of various modes of transportation, such as cars, public transit, bicycles, and pedestrian walkways, to create a seamless, efficient, and safe transportation system.

Scenarios

Below we cover some key scenarios that are common in many connected mobility implementations:

Vehicle and user provisioning

Vehicle Provisioning links the Telematics Control Unit (TCU) with the Vehicle Identification Number (VIN). Vehicle Provisioning also enables secure and automatic provisioning of security certificates and installation of latest firmware. Such activities are performed before the vehicle leaves the factory or upon swapping of the TCU. End of Line (EOL) processes involve configuration and validation of the Telematics Control Unit which includes: setup the certificate for identity of the vehicle, provision the SIM with the Mobile Network Operator, and set the state of the vehicle in the Connected Mobility Platform. User Provisioning involves activating the Connected Mobility services for the customer in the Customer Relationship Management (CRM) and the Billing systems, and activate the SIM in the Mobile Network Operator (MNO) systems which allows the customer to actually use the service.

• Vehicle provisioning reference architecture

• User provisioning reference architecture

Vehicle connectivity management

Vehicle connectivity management helps ensure resilient, secure, bidirectional connection between the vehicle and the cloud. It enables high throughput data transfer, low latency events and supports communication across all devices, in both low latency local area and wide area network connection.

• Vehicle connectivity management reference architecture

Vehicle data management and insights

Connected vehicles have hundreds of controllers and sensors producing thousands of individual data elements for operating, and conveying the state of a vehicle. Vehicle data management helps vehicle manufacturers to harness data as an asset, to drive sustained innovation and create actionable insights and improve their customer experience. Vehicle manufacturers are seeking cost-effective ways to simplify the process of collecting data from vehicles that are connected to the cloud help power insights and improve vehicle performance while maintaining the highest levels of confidentiality and security.

• Vehicle data management and insights reference architecture

Connected mobility core services

Vehicle manufacturers can deliver value-added services to fleet operators and vehicle operators that helps them improve the vehicle operating experience, such as remote lock or unlock, remote vehicle monitoring, usage-based insurance, and improve experience throughout the vehicle lifecycle.

• Connected mobility core services reference architecture

Manufacturing Industry

---

Introduction

In the Manufacturing Industry, the workloads vary quite a lot. It is not possible to address each of these workloads, so this documentation will focus on a couple of popular ones. For other AWS services that you may use within your architecture please see the IDR Alarm Best Practices - AWS Services sections.

Common Manufacturing Workloads

Below we cover some common workloads within the Manufacturing industry:

Product Defect Detection

Below is a Serverless Architecture for Product Defect Detection Using Computer Vision. This architecture can be used for camera-based in-line or end-of-line quality inspection. Supports automated or one-time anomaly detection using image classification in the cloud; real-time monitoring and notifications; and analytics and insights from the classification results.

• Product defect detection using computer vision reference architecture

Anomaly Detection

Below is an architecture which detects anomalies for industrial workloads. It uses IoT, analytics, and machine learning services to inform operational technology teams of performance anomalies.

• Anomaly detection reference architecture

Recommended Metrics to Monitor for both the Automotive and Manufacturing Industries

We recommend using the below metrics to create and configure alarms based on the above sample architectures and advise to follow the Practices for Observability from the AWS Well-Architected, Operational Excellence Pillar located here