CHALLENGE
A global provider of urban mobility services needed to make it easier for its after-sales teams to find the information they required. Trying to identify relevant data among volumes of legacy documents and unstructured content was an increasingly slow and difficult task, making it impossible to fully analyze customers’ requests for repairs. Field operators were struggling to access the knowledge for individual jobs, while new parts were being sent out even when they were not needed.
SOLUTION
The company used data science solutions on the 3DEXPERIENCE platform to store and structure its legacy technical documentation and make it easier to navigate. Cutting-edge generative AI was also used to create a virtual companion, which provides a simple, intuitive way for technicians to locate the information they need.
BENEFITS
The solution structures legacy documentation to make it searchable and contextualized according to users’ needs. It empowers users to assess customers’ service requests and assign actions and new parts accordingly. Field maintenance technicians can resolve problems faster, without wasting time searching through volumes of technical documents. The AI-powered virtual companion provides natural-language answers to their questions, with direct links to the relevant document and paragraph.
Too Much Information
More than a century of successful growth brings a stunning professional pedigree – along with volumes of legacy data. Efficient service is built on intelligence, and that is difficult to deliver when you’re struggling to find the knowledge and context you need amid decades of documents and databases.
A leading global provider of urban mobility services faced exactly this challenge. The company receives up to 2,000 field maintenance requests each month, but its existing system left it handling them on a purely reactive basis. To analyze the root cause of each issue and identify the right solution, teams had to wade through legacy technical information contained in multiple databases and millions of documents, most of which were text-based. Timely analysis was impossible, so new parts would be sent, whatever the problem and field operators would arrive without a full plan of action to fix the equipment.
To enhance the quality of its field operations, the company wanted to simplify its knowledge and put it into context. It needed a fast and simple way for field operators to locate relevant documents and find the knowledge they required within each document.
The company chose the Dassault Systèmes 3DEXPERIENCE platform, including data intelligence solutions, to store and structure its legacy technical documentation and make it easier to navigate. Cutting-edge artificial intelligence (AI) was also used to deliver a simple, intuitive way for technicians to locate the information they need.
Unlocking Stored Knowledge
3DEXPERIENCE provides advanced data science capabilities to deliver knowledge from documents in a targeted, real-world context. That knowledge can then be laid onto a virtual twin – a working 3D model of the real-life situation that can be accessed anywhere, on any device through the 3DEXPERIENCE platform. As a result, each user can see exactly how the information relates to the job they’re working on, and they can share that information securely with anyone who needs it.
Notably, Field Quality Intelligence solution supports better decision-making by unlocking the knowledge stored in legacy information systems so it can be applied to current cases. By using AI to identify resolution reports for similar situations that happened in the past, it delivers insights that help technical support specialists to efficiently pinpoint root causes relating to the problem they are investigating.
These were the building blocks for a system that would connect the company’s people with the contextualized information they need to ensure safe, high quality and accurate installation and maintenance services.
Configuring Searchable Content
The first step was to take the company’s unstructured, legacy content and create a structured way for users to find the right document – and the right information within that document – for the job in hand.
Storing all the technical documentation for the company’s product lines, safety and other practices in the 3DEXPERIENCE platform made it possible to establish an enterprise taxonomy model that would make it easy to search.
Taking PDFs as an example, each document was split into paragraphs. These paragraphs were then imported into the 3DEXPERIENCE Text Intelligence Studio. The company’s own vocabulary was applied, to standardize the solution on common technical wording used across the enterprise.
The system was configured to extract standardized information from the content and create taxonomies to classify it. This yielded a list of categories for each document and paragraph, which correspond to groups of keywords. Making the categories available in the semantic data lake also made it possible to enrich the content, cross multisource and filter the data.
The Convenience of Virtual Companion
Simple, intuitive and accessible tools are at the heart of efficient service, so the company wanted a more spontaneous way for its field operators to find information. For example, if people could get intelligent answers to their questions using natural language, they would be able to resolve technical issues much faster.
This was made possible using an AI agent, called virtual companion, which was developed based on generative AI. They copied the Text Intelligence Studio project, connected a chatbot to it and trained it by asking questions.
The virtual companion builds answers to field operators’ questions from legacy documents. It gives natural-language answers that are easy to understand, with direct references to the documents and paragraphs it has drawn on. This provides a clear path for users to check the details and dive deeper into the relevant knowledge.
Superior Service, Faster Resolution
The solution helps the company’s maintenance organization to swiftly assess customers’ service requests relating to maintenance, repair or defects so that parts and field workers can be assigned appropriately. And out in the field, technicians can quickly find the knowledge they need in the context of their work, without spending time searching through documentation. As a result, it is helping to improve the speed and quality of the company’s field maintenance activities.
For example, when a customer calls to report damage to its equipment, the call center operative sends a technician to assess the damage. By asking the AI-powered virtual companion how to proceed, the technician can search all the relevant documentation and receive a natural-language answer with all the sources cited. This means they can swiftly identify the best way to fix the problem, including which parts to replace.
Elsewhere, a field maintenance technician might call Technical Support to refresh his memory about the processes, tools and safety equipment needed for a specific job. A search by the support team member rapidly points to the right sections of the relevant documents, which the support team member can deliver to the technician’s mobile app with the click of a button.
People rely on the company’s equipment to access amenities, goods and services, so it’s essential that teams can efficiently assess and resolve any problems that occur. The 3DEXPERIENCE platform, has delivered a smart, AI-powered solution to enhance investigation, streamline field maintenance activities and ensure the quality, safety and speed of issue resolution.
Morgan Zimmermann, CEO of Dassault Systèmes NETVIBES
Download the eBook to discover more data science solutions in action!
