How generative AI is revolutionizing the way we find and procure industrial parts and procedures

Any business that deals with industrial parts—OEM or aftermarket—knows the routine. Searching, matching, and retrieving information on a particular part or several related parts is a process that can take hours, if not days. If you need to produce a quote for a bill of materials that includes 1,000 line items, those days can turn into weeks. Meanwhile, you run the risk that your customer loses patience and goes in another direction.

Really knowledgeable professionals could speed up this process, but they’re often in short supply. Some technology solutions can help, but typical lookup tools and databases tend to multiply over time, introducing complexity via incompatible taxonomies and item numbers.

The good news? Artificial intelligence (AI) and machine learning have evolved to offer a third option for industrial parts management—one that doesn’t have to involve replacing the systems that are already in place. What does this kind of intelligent search look like? Let’s find out.

Harnessing the power of artificial intelligence

Generative AI tools are now ubiquitous, with people using them to write essays, jokes, computer code, and even music. Generative AI uses large language models, or algorithms, that can understand, process, and generate both structured and (in some cases) unstructured data. Meanwhile, the technology learns patterns and relationships between different data points to produce more relevant insights in less time.

Unstructured data is common for industrial operations. Imagine you’re a customer looking for a flange made of either steel or brass and you also have related requirements in dimension and pressure ratings. To find the right part through a keyword search, you would need to know the exact part number—which can vary by distributor and manufacturer—or speak “live” with a salesperson who may not be familiar with the part.

Procedures can be even more burdensome. Suppose you need to know what the sulfur hexafluoride (SF6) fill process is for the 345kV SF6 circuit breaker on a transformer at an electric substation. (If you’re not in the energy business, trust that we’re not making this up.) To get an answer under the usual circumstances, you might need to:

• Drive to the regulator station;
• Open the breaker cabinet and pull out the instruction manual;
• Find the section about the SF6 fill procedure;
• Pull out the SF6 fill fittings case; and
• Match the fittings in the case to the instructions and the fittings connections on the breaker.

Nobody loves having to do this, which is why Deloitte launched Industrial Search^™, putting AI to work for operations.

Streamline operations with the power of AI

Artificial intelligence learns by applying what humans train it to do. Neural networks power the AI in Industrial Search™, trained on large datasets to learn patterns and relationships between different data, mimicking how the human brain works.

Search functions that use this kind of technology speak the same language as humans. Instead of customers having to research a replacement oil gauge themselves, for instance, they might simply ask your system: “What’s the part number for the Cops tank oil gauge for the transformer at a specified electric substation number?” If they’re not sure how to interpret a schematic or make the replacement, your AI-powered Industrial Search^™ solution can assist by providing the specific information they need in the moment.

The capabilities don’t stop there. Say you or the customer identified a missing or broken part. You need to know the material cost to replace it. For infrastructure construction projects, this ordinarily means developing quotes on large bills of materials (BOMs). You’ll then need to research each item, identify the part vendors, quote cost and lead times, and compare the results.

With AI-powered Industrial Search^™, this process becomes much simpler. The system can identify compatible units accurately and timely even if the customer has limited or imperfect information. From there, you can develop a BOM, return product matches, and select the right ones based on vendor, cost, and other requirements. Put in the order and your customer is on their way.

Transform your organization and meet customer needs

Traditional search doesn’t necessarily differentiate between multiple meanings of the same word. Neither does it always deal with synonyms and abbreviations well. And you can’t expect it to close gaps when it isn’t able to contextualize and understand relationships.

AI-powered Industrial Search^™ can do all this and more, helping your organization to:

• Boost productivity;
• Avoid cost of poor quality;
• Slash inventory management costs; and
• Accelerate work in the field.

In our second blog, we will explore ways to get started on realizing these types of benefits and illustrate them with real-world examples. In the meantime, if you have any questions or would like a briefing on how Deloitte’s Industrial Search^™ solution can work for your organization, please contact our team.