Remote Vehicle Diagnostics with AWS IoT FleetWise and Amazon Connect

Remote vehicle diagnostics is the ability to remotely detect and diagnose vehicle malfunctions. Connected mobility and improvements in On-Board Diagnostics (OBD) have made it possible to remotely diagnose vehicle malfunctions. Prior to connected vehicles the repair process was reactive. Service centers had little insight into vehicle malfunctions until they were brought in and inspected. Service technicians spent valuable time diagnosing problems, creating estimates, and ordering parts before repairs were made. Mean Time to Repair (MTTR) is affected by time waiting for inspection, work authorization, and parts. With remote vehicle diagnostics, these processes can be done before the vehicle arrives at the shop for repair. Utilizing this capability is important for service centers and fleet operators to keep up with maintenance and repair of the growing number of vehicles.

Service centers and fleet operators face challenges to implement a scalable remote vehicle diagnostics solution. The solution must collect only the data necessary for remotely diagnosing the problem to keep costs low. The solution must engage with the customer once the issue is detected and support omnichannel communication to keep the customer engaged. Finally, the solution must keep track of the repair throughout the repair workflow and provide integration with other systems.

In this blog, we will cover a remote vehicle diagnostics solution that changes the repair workflow from being reactive to proactive. This is done by remotely detecting vehicle malfunctions in near real-time. The solution then creates a service case to track the vehicle repair. It uses diagnostic trouble codes (DTCs) and sensor data to remotely diagnose the vehicle malfunction. The vehicle owner is notified of the issue via SMS text and asked to schedule an appointment. When the owner contacts the service center, they are automatically linked to the case, which is presented to the service advisor. The owner can add additional details at that time. Once the repair is complete, the owner is notified that their vehicle is ready.

Overview of the solution

The solution integrates multiple serverless AWS Services to provide an end-to-end implementation for remote vehicle diagnostics. Let’s review how each AWS service integrates into the remote vehicle diagnostics solution, shown in Figure 1.

• Automotive manufacturers and fleet operators build modern connected vehicles using AWS IoT Core and AWS IoT FleetWise that provides near real-time telemetry data.
• Data is stored in Amazon Timestream and Amazon DynamoDB allowing the solution to remain performant at scale.
• You can use AWS Step Functions to build workflows that integrate services. These workflows provide powerful automation capabilities that can be expanded to include other systems such as parts inventory or cost estimate systems.
• Amazon Connect is an omnichannel cloud contact center. It tracks and manage the repair through multiple interactions between the customer, service advisor, and service technicians. It also delivers information and guidance to the service advisor to increase productivity.
• Amazon QuickSight is a business intelligence service that is used to visualize vehicle telemetry data. QuickSight uses Amazon Athena to query DynamoDB for vehicle and owner information.
• Amazon EventBridge automates workflows from system events such as updating the repair case.
• Amazon Pinpoint is an omnichannel customer engagement service that provides personalized and targeted messaging to the customer. It sends notifications and updates to the customer about the repair.

Figure 1: Solution Architecture

Walkthrough

Let’s explore the architecture step by step.

1. Configure and deploy the AWS IoT device software and AWS IoT FleetWise Edge Agent on the vehicle gateway or TCU.

2. Configure AWS IoT FleetWise, including vehicle models, decoder manifests, vehicles, and campaigns to collect vehicle telemetry data. Campaigns are then setup for each DTC to only collect sensor data related to that DTC. This helps keep costs low.

3. When a DTC is activated, that will start the collection of telemetry data of the campaign configured for that DTC.

4. The telemetry data for the campaign is collected and stored in Timestream.

5. The vehicle software will then use AWS IoT Core to update the vehicle device shadow with the DTC. An AWS IoT rule action calls the AWS Step Functions workflow when the device shadow DTC value is updated.

6. The vehicle service Step Functions workflow, shown in Figure 2, performs the following actions. First, it gets the vehicle and owner information from DynamoDB. Then, it checks if an active case exists for this vehicle based on the DTC to prevent duplication. Next, it creates the Amazon Connect Case for the vehicle issue with all the relevant information and attaches the customer profile. Finally, Amazon PinPoint sends an SMS text to the customer with the case details and instructions on how to schedule an appointment.

Figure 2: AWS Step Functions Workflow for Vehicle Service

7. An available service advisor in the Amazon Connect contact center will pick up the case and review the issue by clicking the diagnostics dashboard link. Here, the service advisor can update the case notes, check parts inventory, or generate cost estimates for the customer. Amazon Connect Cases makes it easy to create, collaborate on, and quickly resolve customer issues that require several customer conversations and follow-up tasks. Custom case fields are added to track industry specific fields such as the VIN, model, and year of the vehicle. The custom case fields are then added to a case template to track the vehicle repair.

Figure 3: Amazon Connect Case

8. The QuickSight diagnostic dashboard queries sensor data from Timestream and DTC information from DynamoDB. Let’s walk through an example on the dashboard to see how the dashboard helps service advisors and technicians with troubleshooting the vehicle malfunction before the vehicle arrives at the shop.

A. The DTC shows the affected vehicle system.

B. The fan relay is in the running state. This tells the service advisor that the relay is not the issue.

C. The problem isn’t due to a faulty sensor as the coolant temperature is rising.

D. The coolant fan voltage, however, has dropped below the operational threshold. This may be due to defective wiring, a defective fan, or defective power supply. The issue has been narrowed down to a problem with the coolant fan voltage, with additional sensor data we can narrow it down further.

Figure 4: Quicksight Diagnostic Dashboard

9. The owner receives an SMS text message notifying them of the malfunction. The text includes the case reference and details to contact the service center. The owner can now schedule an appointment.

10. The owner clicks the service number from the text message to call the service center. The customer is greeted by name. The Interactive Voice Response asks the customer if they are calling about the recently created case. The customer confirms that they are calling about the case which automatically links the call to the case. The customer follows the prompts to connect to a service advisor. The service advisor confirms the vehicle issue and assists the customer with scheduling an appointment.

11. When a service advisor makes an update to a case, the update event is sent to EventBridge. An EventBridge rule calls the service update Step Functions workflow. The workflow gets the owner information and sends an SMS text update to the owner notifying them of the update. Once the vehicle repair is complete the owner will automatically be notified that their vehicle is ready for pickup.

Figure 5: Example SMS text

Conclusion

The solution shows how AWS IoT FleetWise can be used to remotely diagnose vehicle malfunctions. It also shows how Amazon Connect and Amazon Pinpoint can be used for case management and enhanced customer engagement. Using sensor data to remotely diagnose vehicle malfunctions improves the efficiency of the repair workflow by allowing time consuming activities to be done before the vehicle arrives at the service center improving MTTR metrics. These activities include the time spent by technicians diagnosing problems, ordering parts, and getting repair approval from the customer.

Service centers and fleet operators can add additional value to this solution with additional sensor information and system integrations for more efficiency. Analyzing more sensor data and past repair data with machine learning algorithms can automate repair diagnoses. Integrating inventory and repair estimate systems can automate cost and time estimates. Online appointment scheduling can be integrated to make scheduling easier for the customer. Amazon Pinpoint can send targeted marketing campaigns to customers when a condition is met. For instance, a tire discount email can be sent when the vehicle odometer exceeds 50,000 miles. The solution provides a flexible workflow that can easily be expanded to provide more automation and better customer engagement.

For more information visit AWS for Automotive to learn how AWS empowers digital transformation for automotive companies.