AWS for Industries
Automotive Call Center Services Solution Using Amazon Connect by WirelessCar
The number of connected cars on the road is set to rise from 125 million in 2020 to over 470 million by 2025. The push for electric vehicles is helping to accelerate a new era of modern vehicle architectures that are more centralized, with fewer distributed electronic control units driven by the introduction of domain controllers. We are also seeing further centralization into zonal architectures, which ultimately leads to reduced cabling, weight, and complexity. This shift from hardware to software will only increase customer demand for more customized in-vehicle experiences. Amazon Web Services (AWS) is helping provide undifferentiated services, solutions, and support to help automakers become software-driven companies, and AWS Partner WirelessCar is running connected vehicle services on AWS.
By July 2021, WirelessCar—which has been working since 1999 to create connected vehicle solutions for cars, commercial vehicles, and fleet management services—had connected more than eight million vehicles and counting in over 75 countries. This enables customers, such as Volkswagen, Jaguar Land Rover, Daimler, Nissan, and Volvo Cars, to use the full value of connected services on AWS. Among connected vehicle services, call center services for emergency/SOS, breakdown, service, and concierge/convenience are becoming essential for the brand customer experience and customer loyalty to original equipment manufacturers (OEMs). This article provides more details.
WirelessCar developed a solution that automates the contact flow, or a customer’s experience with a contact center from start to finish. The solution uses Amazon Connect, an easy-to-use omnichannel cloud contact center, and Machine Learning on AWS, a set of machine learning services and supporting cloud infrastructure that puts machine learning in the hands of every developer, data scientist, and expert practitioner. By using these services, WirelessCar has increased customer satisfaction and loyalty, reduced OEM costs, and provided OEMs with qualitative insights. This blog post will focus on emergency and breakdown services.
Connected vehicles enable essential data to be shared directly with the automaker to streamline the customer experience when it comes to roadside assistance, stolen vehicle recovery, or emergency response. Equally important is that this customer-to-OEM interaction can largely be automated to create 24/7 response time. Using connected car data to provide these essential services provides a tremendous opportunity for automakers to deliver services often outsourced to third parties.
Call services for connected vehicles create a direct relation between the driver and the vehicle OEM. These services, such as roadside assistance (RSA) or stolen vehicle assistance, are becoming increasingly challenging due to the growing complexity of vehicle systems, but these developments present a lot of potential as well. Vehicle OEMs need to find a way to tackle the issues that arise and do so with professionalism. If successful, they will enhance consumers’ brand perception and increase profits. If not, consumers will likely use other services to solve their issues and will reward competitors instead. Great call services—supported by highly trained staff who are experts on resolving the specific customer issue efficiently—can reinforce the OEM’s brand values and live up to customer expectations. However, this requires a network of call center partners across national borders, and vehicle OEMs must be able to manage these with flexibility and cost efficiency.
The following are reasons why OEMs should adopt this blog post’s proposed solution:
- Regulations and Customer Safety
According to the European regulation EU 27: 112, it is mandatory to have an emergency call service for new vehicles. Currently, calls are made to a staffed call center in which the operator must determine the severity of the situation by speaking to the driver. A solution that uses vehicle data can automate the call to the emergency response center, such as when it detects a hard impact or airbag deployment. Critical details like the speed at impact, numbers of airbags deployed, vehicle operation status, and even video footage can instantly be transmitted to the operator. Based on predetermined criteria, emergency services can be dispatched even if the driver is unresponsive.
- Customer Experience Improvement
To provide rich experience in the car, lifelike conversational services for vehicle breakdown, emergencies, vehicle-related questions and notifications, and concierge services are becoming more popular. These services provide customers with convenience and a feeling of individual service, and they increase customer loyalty. The natural language processing (NLP) and transcription possibilities of Amazon Connect will help to generate qualitative insights. OEMs can use speech analytics powered by machine learning to access live call transcripts, understand customer sentiment, and identify call reasons in near-real time, and identify trends like common parts failures before they escalate to a broader recall. Contact centers can also configure rules to generate tasks in near-real time for issues to be routed to consumer affairs or other appropriate teams. For convenience and concierge services, offerings like Amazon Connect, Amazon Machine Learning solutions for NLP, and Alexa Skills Kit SDKs—software development tools and libraries that provide programmatic access to Alexa features—can automate the contact flow. For example, an electric vehicle driver can use voice commands to learn the best route that will provide minimum charging time while enabling the driver to enjoy their favorite food while they wait.
Architecture and Solution
WirelessCar’s solution for call center services creates the right conditions for vehicle OEMs to offer call center services for emergency assistance, RSA, stolen vehicle assistance, and driver assistance. The solution runs on AWS and provides portals and APIs to facilitate a smooth integration between vehicles and call centers. Due to a standardized approach for connecting vehicles with call centers, integration efforts are reduced. This presents businesses with opportunities for greater agility when they are selecting the right partners for setting up call center networks that can successfully support a broad service portfolio with global coverage.
In this solution, in emergency or breakdown situations, a call will be automatically triggered from the car, along with a snapshot of car and user information that is sent to the connected car backend. The customer care agent receives the call and the user and car information through the WirelessCar user interface (UI) so that they can provide direct help to the user. A user will contact a virtual call center or Amazon Connect instead of a call center agent first. And depending on the scenario, the contact flow might be automatically launched to meet the user’s needs. Scenarios are explained in the following use cases section. WirelessCar provides APIs that enable call centers to retrieve necessary information about the car or user in the UI. WirelessCar uses APIs alongside Amazon Connect to create an omnichannel cloud contact center that provides automated, scalable, and intelligent contact flows to deliver contact center services easily and cost effectively.
In each car, the call options of emergency/SOS (eCall), breakdown (bCall), service (hCall), and concierge/convenience (cCall) are available in different ways (as shown in Figure 1: Car Call Options).
If these buttons are pushed, if airbags are opened, or if other sensors detect an emergency situation, the snapshot of car sensor data and events is sent to the WirelessCar backend. At the same time, eCall is triggered from the car (as shown in Figure 3: Architecture Diagram). Each type of call service from a car has a distinct telephone number. Based on the number, Amazon Connect detects the type of call and starts the automatic contact flow for the call. The solution will then assign an incident case number for the call and put it in queue to be attended by the call center agent. At the same time, the WirelessCar backend will match the incident case number with the snapshot of data received from the call. The car snapshot has details about the sensor data, car information, location data, and user data. The user, car, and incident information are displayed automatically on the screen of the call center agent (as shown in Figure 2: Call Center UI). The call center UI helps with case management, start of remote services (like door opening, engine starting, and more), reports and case notes for documentation and audits, and location information in maps. This UI application is translated in more than 15 languages. After talking to users, the call center agent can take necessary actions, like dispatching the nearest emergency service teams.
In certain events, the car calls the Amazon Connect number. The customer hears the acknowledgement of the call and confirmation of the purpose of the call. The contact flow automates Amazon Connect customer calls, and in an emergency, the call will be forwarded to the call agent. In the case of a breakdown call, it will be forwarded to the call center agent or RSA services directly. Through Contact Lens for Amazon Connect—a feature that helps companies better understand the sentiment and trends of customer conversations—companies can configure which calls they want to analyze in the Amazon Connect contact flow by selecting the speech analytics check box in the “set recording and analytics behavior” block. Contact Lens also enables customers to redact sensitive information from the call transcripts. They can enable or disable this option by choosing “redact sensitive data” in the same contact flow block. Once they have completed this configuration, Contact Lens will automatically start analyzing calls that pass through this contact flow block. The call recording and analysis are saved in Amazon Simple Storage Service (Amazon S3)—an object storage service that offers industry-leading scalability, data availability, security, and performance—in an Amazon S3 bucket. This solution also does the postcall analysis using Amazon Comprehend—an NLP service that uses machine learning to uncover information in unstructured data—to gain further insights about the customer call. For monitoring, quality control, and trend analysis, all these insights can be visualized on Amazon QuickSight, a scalable, serverless, embeddable machine learning–powered business intelligence service. Contact flow for service and concierge calls can work in this same setup.
Figure 4 shows the execution of the contact flow for a bCall, or breakdown call, such as when there is a flat tire and no injured person. In this case, with the help of the car sensor snapshot data, the solution can determine and confirm the reasons for the breakdown by talking directly with the user. Depending on the context, Amazon Connect can talk to the customer by asking questions, using Amazon Lex—a service for building conversational interfaces into any application using voice and text—to understand the natural language context, and it can offer different options to the driver and also take actions like dispatching the RSA at the vehicle location. This entire process is automated and documented using an Amazon Connect contact flow. Amazon Connect includes a set of default contact flows, so customers can quickly set up and run a contact center. However, customers might want to create custom contact flows for their specific scenario. As a user interacts with the contact flow, this solution enables them at any time to request to transfer the call to a human agent. The agent can see the history of conversation and pick up the conversation where it was left by the contact flow. This solution can enhance customer experiences by providing a function for customer care and by directing vehicles to preferred service partners (licensed dealers). The solution also integrates with RSA providers and shares any relevant data for assistance. WirelessCar works with a number of RSA partners. Once the RSA is dispatched, the user gets the contact information and near-real-time tracking information on the map about the dispatched service. This makes the user feel secure and confident, as well as loyal to the car brand.
Vehicle accident and emergency situations are recognized by this solution using the car’s sensor data, like data on the deployment of the airbags, vehicle speed, or diagnostic errors. According to EU 27: 112, it is mandatory for OEMs to provide emergency call center services. In the case of an emergency, an automatic call is triggered from the car, and a snapshot of the user data and car diagnostics are sent to the car backend. A call center agent can judge the severity of the accident based on the data and talk directly with the user. WirelessCar is working toward implementing a video call service that can use a dash cam in the car, with the user’s permission, in emergencies. Figure 5 shows an example of the contact flow for an eCall.
- Low integration efforts: a standardized way to integrate global call center services that will reduce time to market and lower integration and maintenance costs
- Streamlining processes: streamlined and standardized processes across call centers by providing a web-based client for agents that will increase efficiency and reduce training costs
- Reliable and secure: a secure solution that protects the vehicle and personal data, operates with several million connected vehicles across several markets, and saves lives every day
- Better customer experience: enriches services with additional data, providing agents with the means to deliver a better, contextualized end-customer experience
OEMs can use Alexa SDK, Amazon Connect, and Machine Learning on AWS to automate the call center contact flows for eCall, bCall, hCall, and cCall. These contact flows can be integrated with connected vehicle systems to increase customer satisfaction and loyalty, reduce OEM costs, and provide OEMs with qualitative insights. Contact us to help you realize this solution with WirelessCar.