Using AWS to seamlessly and cost-effectively scale localization needs for automotive customers

Manufacturers are required to create health and safety documentation to protect the health and well-being of users, passengers, mechanics, and employees. Every vehicle requires the production of safety documentation, which can be considerable in terms of the number of volumes, the size, and the complexity. For service manuals, this documentation can include topics that range from the operation of complicated machinery to the initial setup of new computer hardware.

Localization is a process to adapt content, products, services, and materials for international marketplaces, which feature differing languages, regulations, and cultures. Translation is a core part of localization; however, doing it manually and using a human-centered approach can be resource intensive and expensive.

If localized documentation is produced to an unsatisfactory standard or is not timely, the outcomes are potentially serious and must be managed proactively to reduce any risks and ensure the safety of any individual associated with a vehicle.

In this blog, we will detail how an automated localization solution powered by Amazon Web Services (AWS) and built by Valcon has made it possible for a global automotive manufacturer to streamline its localization operations and improve the efficiency of translating its vast libraries of technical content into multiple languages.

Valcon has a specialist team of over 800 professionals focused exclusively on business transformation through the effective use of data. With operations in the United Kingdom, Benelux, and the Nordics, the company is an AWS Advanced Tier Consulting Partner focusing on data science, strategy, governance, management, and analytics.

Importance of cloud-based localization and translation of manuals for this innovative automotive manufacturer

For this automotive manufacturer, the process of localizing content was time dependent and not cost effective. It relied on labor-intensive human translation. With so much content that needs translating on a continual basis, costs were rising while efficiency levels would fluctuate, causing a significant impact across the business. Using cloud-based automation for the translation process was the next logical step, and to facilitate this, the customer used the expertise of Valcon to help realize its goal.

After understanding the objectives of the automotive original equipment manufacturer (OEM) and the challenges it faced, it became clear that providing the OEM with a dependable and efficient automated solution was the goal. Valcon designed and created the Translation Data Pipeline Service using innovative AWS technologies. Using the highly effective automated service, the automotive OEM can now localize and distribute its content for global marketplaces much more quickly and cost effectively than ever before.

Container-based localization solution

As a member of the AWS Partner Network (APN), Valcon is well versed with the cloud service capabilities of AWS. Valcon was able to design, create, and deliver a working solution in under 3 months with any future updates delivered through the cloud-based managed service.

The following diagram shows the various services being used for the localization solution:

Figure 1: Valcon localization architecture

The localization workflow starts with the OEM’s technical writing and engineering teams, which create the initial documentation. The documentation identified as requiring specific localization is securely uploaded to the cloud using the AWS Transfer Family—which securely enables recurring business-to-business file transfers. AWS Transfer Family stores the data using Amazon Simple Storage Service (Amazon S3), an object storage service offering industry-leading scalability, data availability, security, and performance. Then, Amazon Simple Workflow Service (Amazon SWF) is initiated and manages the entire workflow. Amazon SWF lets developers build, run, and scale background jobs that have parallel or sequential steps. From there, the E2E workflow containers are orchestrated and scaled using Amazon Elastic Container Service (Amazon ECS), a fully managed container orchestration service, to determine what compute resources are required to complete the job.

Additional resources are scaled up and launched using other third-party translation services from XTM International and CrossLang, which also run in containerized workloads managed by Amazon ECS. XTM International delivers the multilingual translation management while CrossLang provides machine translation services. Throughout the process, Amazon DocumentDB (with MongoDB compatibility)—a scalable, highly durable, and fully managed database service—is used as a MongoDB-compatible central document database for access from various stages of the service. AWS Lambda, a serverless, event-driven compute service, is responsible for event-driven initiations as well as update notifications. If required, the E2E dashboard can support manual inspection and finalization of the produced documents. Finally, the localized documents are provided through Amazon S3 for the user to download and securely distribute to various teams around the world.

Valcon designed and built the Translation Data Pipeline Service using innovative technologies. It also built and maintained the service in the cloud using a full DevOps pipeline that relies on AWS DevOps services. This makes it possible for future updates to be deployed quickly, adding new features so global teams benefit from new functionality and efficiencies.

Value creation and trust building with the automotive customer

By using AWS, Valcon is able to facilitate seamless global scalability and cost savings. The AWS infrastructure scales dynamically to meet the varying needs so that customers only pay for the infrastructure consumed on demand, with no upfront costs or long-term contracts.

The combination of accuracy, speed, and scalability provide high levels of efficiency, resulting in significantly reduced costs when compared with human-centered translations. The ability to recognize and remove duplicate content is a standout feature, with millions of text passages and images deduplicated over a 3-year period. As the service is constantly learning, translations are becoming more efficient over time.

“Had nothing changed, the customer’s translation budget was forecast to go from less than €5 million per annum to more than €25 million over a 5-year period. The service has translated well over one billion words so far, which is the equivalent of a small library.” —Kenny Lau, cloud architect, Valcon

The critical need for quality

The service is able to translate complex documentation for different marketplaces and regions efficiently and reliably. Accuracy and dependability engender trust among the key stakeholders while the centralized monitoring of key performance indicators and operational data help drive data-led decisions and improvements to business processes. The highly reliable uptime of the translation services help facilitate compliance while de-risking compliance regulations.

High-quality translations are facilitated through nearly continuous machine learning and the ability for humans to override and correct issues if necessary, offering a comprehensive safety net. Quality assurance checks are built within the automated service, which provides translated materials that are reliable and ready for publication for marketplaces around the world.

Metrics of success for Valcon’s automotive localization solution

• 70% time-to-market reduction
• 67% average translation cost reduction
• 5% translation error rate (reduction from 10%)
• Over 30 million units of duplicate content identified and managed

Speak with Valcon to learn more about how you can automate translation and localization services to improve quality and reduce cost.

Related resources:
AWS automotive solutions