How To Optimize Supply Chain Efficiencies With AI

To thrive in today’s landscape and maintain a competitive edge, CPGs must optimize their end-to-end operations. This means tackling any inefficiencies at the supply chain level to reduce unnecessary costs related to burdened legacy systems and elevating consumer experiences by getting to market faster.In this webinar transcript, Evan Sax, NETVIBES senior sales expert, Dassault Systèmes, provides CGT with a product demonstration on how CPGs can address the biggest pain points that are driving up costs and causing delays. Explore how building in consolidated data structures can reduce or eliminate unattended part data debt, which is often a hit to the bottom line.Lisa Johnston: Hi, everyone. Welcome to CPG/Home & Life Smart Chains: Cutting Costs and Optimizing Supply Chain Efficiencies With AI. My name is Lisa Johnson. I'm the editor in chief of CGT. I'm excited to welcome you here today. We have Evan Sax, NETVIBES senior sales expert, Dassault Systèmes, to review some of the biggest supply chain management issues that are mucking up the works and adding to costs. So Evan has prepared a video for us that walks us through some of today's challenges including the role of unattended part data debt. He's then going to show us a quick demo of their platform. After the video, Evan and I are going to return here for a live Q&A to take any questions you might have. So when you're watching the video, feel free to just drop them right into Q&A as it plays and we'll tackle them as soon as it's over. With that, let's get the video started.Evan Sax: Thank you for the introduction. Hi, everyone. I'm Evan Sacks NETVIBES senior sales expert. And I'm excited to discuss with you today a unique solution around part process optimization and supply chain cost savings. I suspect you may be asking yourself how the two topics are connected and why should I care. Let me offer two scenarios to kick things off. Do your design teams put any real energy into trying to find and reuse that perfect part or container for your next product because they don't know where to start looking? So instead they just design another one that needs to go through an entirely new purchase request test and validation process. Again. Or perhaps your procurement teams are buying similar components at vastly different prices and they are even aren't even aware of it or have no way of easily identifying it. We're going to take a look at how we can leverage AI to efficiently help identify similar or identical parts within our entire library. This improvement in your process can pay big dividends by not only optimizing your engineering and design teams but also driving 3%-plus cost savings out of your supply chain by utilizing a three-phased approach.Our unique solution utilizes AI to do a CAD agnostic 3D geometric comparison of parts within your library to provide visibility into your data that wasn't achievable before we bring your 3D structured and unstructured data together across multiple data sources including multi-CAD and multi-PLM environments, and any other data source. This insight helps design teams find and reuse parts in a matter of minutes. It also creates unique insights for supply chain, providing new perspectives to look at your data and find new opportunities for cost savings and other mitigating factors. Phase one, a multi-source part visibility where an engineering part search is created to help streamline the finding of parts and reuse of existing parts. You can now search for a part in a multitude of ways in a matter of minutes, bring the part into your design, and then continue what you are doing without having to spend the time on designing and validating a new potentially duplicate part — or you can start from a current design that meets most of your requirements and make the last few modifications required instead of starting from scratch. But that's not all. You also introduce an easy way to govern new part introductions in order to reduce the number of new part procurement requests. A small but sizable supply chain cost saving when you add the total cost savings over the year. Phase two — part standardization — is where all similar and duplicate parts within your library are identified. A standard part can then be identified for use going forward. This enables you to start a data cleanse by driving standardization and in turn reduce your part digital debt. Phase three: part process optimization. This is where supply chain savings can be identified across several different scenarios. We enable the user to slice and dice your data from multiple perspectives at the same time to give you the full picture of impact and value within your supply chain phase. One entails four easy steps. Step one, all part data is indexed. This is completely agnostic of data format or source, all CAD file formats can be indexed simultaneously within a multi-CAD, multi-PLM environment. All metadata within the CAD file will be included in the indexation along with the geometry itself. Step two, utilizing a unique 3D signature of each part is created based on its geometry. Step three semantic data from several other sources can be indexed and aggregated to enhance the signature of each part. This includes structured and unstructured data. For example, supply chain-related data such as cost supplier, etc., can all be indexed and aggregated for each part. Each data set being indexed will be added as an attribute to each part for further comparison and analysis. Step four: A parts search is automatically built utilizing this data to easily identify parts and their associated data in a single UI. No matter what systems those data sets sit in throughout your IT landscape, this gives the user a single point of access for any part information they may require geometrically. Similar parts can also be accessed for each individual part; this search tool will allow the user to look at data across other business units and brands to understand what other innovations and best practices in design, and different areas of the organization have achieved and can be brought into the user's design. An advanced search capability is not just limited to your design team. It can also be extended to help your supply chain. An example taken from one of our customers is late supplier mitigation, regardless of how good your product managers are at managing their master schedules. Suppliers run late all the time and it impacts production. This customer recently had a late supplier for a $5 part. When they went to source the part and the quantity needed, it was going to cost them almost $190,000 in increased costs to get the parts ASAP utilizing the search tool. They found that one of their factories in the Czech Republic had these parts sitting on the shelf with no usage assigned to them. They're able to have the parts shipped to Ireland and keep the production going without having to pay the exorbitant price to source and expedite the component from elsewhere. At this point, a fully functioning part search is available for the organization which can be used in a multitude of ways conservatively, we typically see a 20% to 30% reduction in time to look up part related data and a 15% to 20% reduction in new part introductions. A good example of how these savings are being achieved. We can take a look at one of our customers. We have a home and life customer who is currently looking to utilize this solution to help drive synergies across their different brands in some ways, their different brands build very different items. But the parts that make up these different items are in many ways similar. The customer objective is to utilize part search to help their engineers identify a current part quickly that can be reused to drive efficiencies, not just in their engineering departments but as well as in manufacturing and supply chain. Now we move into phase two, this is where we build a decision aid for part standardization, and step five AI-driven clustering of similar and like parts across the entire library of CAD parts takes place. The system will automatically cluster parts based on similar and identical geometries, making it easy for the user to truly see what similar and identical parts they have across all their systems and business units. Again being an agnostic comparison across different CAD files is achievable. Utilizing the raw file format. No conversion to a neutral file format is required in step six. The user can now go into each cluster of similar parts and see a side-by-side comparison of all these components, including all the additional attributes that were aggregated from all their other systems. At this point. The user can choose what part they want to be the standard or preferred part and what parts they want to be alternates. The designation will appear back in the search UI. As we saw earlier, this functionality can be utilized for other use cases as well. We will take a look at some more shortly as we start to look at the data set from different perspectives, we continue to drive further value out of it. As you can see here, customers continue to further drive a reduction in part duplicates as well as an overall new product design. Secondly, in certain areas, we start to see an impact on the supply chain and the ability to drive out some hard-cost bottom-line savings. What we are starting to see is over time. A linear of your data quickly starts to bring diminishing returns. Now take the same data sets but given the ability to slice and dice the data in ways that are meaningful to the user at the same time being multi-perspective creates a larger sense of data visibility. This allows for more efficient actionable insight that brings a lot more value into an organization. Let's take a look at phase three, which will highlight some of this value. What do we mean when we talk about utilizing a multi-perspective data set and the value we can bring? We've talked a lot about how this solution can impact engineering and design teams in multiple ways. We can take this exact same data set and bring different sets of values out of it. Just by changing the lens. Let's now change your perspective from engineering to supply chain data set is the same, but the value the organization gets out of it is different and incremental to what we've discussed already. We've looked at how utilizing AI for 3D geometric comparison across your library mashed together with it. Your other semantic data could bring a lot of value into an organization through steps six and seven. We can help procurement and supply chain identify different areas of cost savings from price discrepancies to inventory utilization to building stronger RFQ packages, etc. We can add a layer of intelligence on top of this data to provide even further insight into your data. Our intelligence tools are built to solve complex problems and work with large complex data sets. In this instance, our intelligence tools can sit on top of this data and start to identify trends and patterns to help procurement and supply chain, not just identify different opportunities but also where they can bring large value. I was at Precision Cas Parks Corporation part of Berkshire Hathaway for 11 years. Most of that in supply chain. My last role there being a global director of supply chain, I was responsible for almost $1 billion and spent. I can tell you I've seen all sorts of supply chain software and nothing comes close to bringing the insight that we can bring or the ability to bring value and impact in such a short amount of time. I recently met with the director of supply chain of a Fortune 500 customer and we walked through what value our AI-driven solution can bring them. They had never seen anything like it before. And the tone of the meeting quickly changed from another AI software solution we don't need to, “This is a tool that can actually bring us some value and solve some of our issues. Some we didn't even know we had.” Let's take a look at an example. Here is an example from an actual customer proof of value that I performed. As you can see, we have two very similar parts: small three-inch aluminum brackets that have similar volumes but widely different prices. One part is almost seven times the cost of the other when they should be the same costing this customer almost $600,000 a year. The best part is these two parts are the exact same supplier. When we presented this to the chief procurement officer, he practically fell out of his chair. We continued to show five more parts in a similar scenario. The six parts altogether totaled $1.75 million in savings opportunity. That's just six part numbers from a small proof of concept. We have seen this same scenario with multiple customers in multiple industries. As you just saw the savings potential from a supply chain perspective can add up quickly. The values here are conservative; the percentages might look small but in dollar value, they can translate to some very interesting numbers for an organization. The ability to bring visibility to engineering data across the organization with a multi-perspective view from a single access point can bring much-needed insight to drive impact by helping to knock down the data silos. Savings from a supply chain perspective are mostly hard cost, bottom dollar savings. We're not done yet. We have one more area of value to review over time. Many organizations tend to migrate data from one system to another or bring on new systems. In addition to what is currently in place, this type of data movement can cause real headaches for users when they are trying to find the piece of information, they are looking for an added benefit of the park search capability is that no matter where the data is coming from or where it is moved to the user can still easily access the information to the user. They may never know that the data was migrated elsewhere as it still appears in the search in the same format as before with only the source changing name. This allows your teams to feel minimal impact while any sort of data migration is taking place. It also helps drive synergies when new data sources are brought in from within other business units or new acquisitions. OK. Let's take a quick look at the tool here and jump right into part data optimization and supply chain savings opportunities. This is where your user can come in. Look at all the AI-enabled clusters of similar and identical parts within the library and start identifying standard parts within a cluster of similar components. Let's take a look at the universe of clusters. So the size of the boxes indicates the amount of similar or identical parts. So the bigger the box, the more similar parts are found within that cluster. So let's jump into it and take a look. Here's where you can tag a part to be a standard part or an alternate part. That's your standard. This is creating an alternate and you can also completely hide parts from the search component if you would like. If you don't want anybody to see that component in the future, but if the part number and the price currently shown are not enough information to identify which part to tag is the standard, we can do a deeper comparison of all their data attributes. So let's choose five to compare. But you can go ahead and choose as many as you'd like to three 45 to prepare. So here you can make a much more informed decision on which parts to identify as standard and which ones are alternates for future design use. As you can see, this can be done within just a few clicks of the mouse just like before the standard alternate, you can also hide the component. So this screen has many use cases that we could talk about. So let's turn our attention to the supply chain for a moment. Supply chains can come in and create their own clusters to start running their analysis to determine price discrepancies, etc. As you can see, he can do a side-by-side comparison of all the parts in a cluster to understand the similarities and differences. This is where supplier names can be seen side-by-side as well as current costs, etc., down here towards the bottom, these price discrepancies become very easy to identify so the procurement team can go action them quickly. There are also several other valuable paths supply chain can take with this data including stronger RFQ packages of similar parts, more effective supplier consolidation, increased usage of excess and obsolete inventory, mitigating part shortages, etc. The list goes on. So these use cases we just discussed fall under phase two and phase three that we reviewed earlier. They are high-impact use cases that can bring a large amount of value to the organization. These use cases are very focused on specific areas within engineering and supply chain. So let's go take a look at the phase one use case where part search is created and can be utilized across the entire organization. Once the tool is installed and the data is indexed, this part search becomes available almost immediately. The value and impact this brings to an organization can be enormous by bringing all parts and their related data together from multiple systems. Your customer can now easily access any engineering data from a single application. So let's take a look at some of its capabilities. We can search for a part using a part number description, etc. So we just go search for a bracket, we can do that. You can also filter on part attributes to narrow down your results. So over here, and this whole list will change based on the attributes that you choose to bring into your index. Another search functionality is the ability to drop in a CAD file and have the system find all its similar and identical parts on the fly. It is now comparing that CAD file across your entire library of components. In just a matter of seconds, we'll have search results that show all the similar or identical components to that CAD file. The blue is the file that was uploaded and then everything below that are the related search results that it found. So the ease of engineering data access will bring large efficiencies to many parts of the organization. By clicking on a search result, you can get access to all the part data and additional attributes and other geometrically similar parts. So for example, parent and child information, as well as any type of drawing or other documentation associated with that part. So we can see all the parent and children information with this assembly going into the bracket itself, we can see all the other attribute information that's coming in from the CAD file as well as from other sources outside the CAD file. So and then you can also get any other type of documentation. If there are any other documents, drawings, specifications, or documents associated with these parts, they could all be brought in under the relations tab, and then PDF and other types of files are all easily accessible. So supply chain can access the data they need to provide to their suppliers and manufacturers can easily access the engineering data. They previously spent hours digging for users from all over the organization and can now access all sorts of engineering and other data in a single easy-to-use UI. So with that, on behalf of Dassault Systèmes, I'd like to thank you for taking the time to attend this webinar and hope you are starting to think of use cases where some or all of what you heard can help you drive more value out of your data. We look forward to helping you immediately at this point. I'd like to open the session up for any questions.Johnston: Thanks for that, Evan. We're going to dive right into the Q&A just as a reminder, feel free to drop your questions right there and we're just gonna tackle them as we go. So Evan, could you talk about some of the most common challenges that you've seen with CPGs when it comes to aggregating data from these different sources that you mentioned ?Sax: In all honesty, when it comes to where we grab ata from, and the way we bring that data in, we've been able to grab data from every source at every customer that they wanted to bring in. So there are multiple ways and multiple options that we can approach to be able to aggregate that data. And we've yet to really run into a data source that we were not able to get what we needed into the tool one way or the other.Johnston: What about security? What role does security play within all of this?Sax: There are many organizations that, you know, depending on who's accessing what data, they really don't want everybody to see everything. And so we can very easily follow all the different types of security rules and rules that are laid out within the organization as each user logs in. They only have access to the data that they can see throughout the organization.Johnston: Can you talk a little bit about some of the different ways that businesses might be able to prepare their teams to successfully adopt new technologies like this?Sax: Sure. So what we're looking at here is really understanding and driving a new or taking a current process and making it far more efficient, right? So the end result of what's trying to be achieved is the same as what was trying to be achieved previously, but now it's just in a new way. So it's looking at your business process and understanding how does this tool come in and either partially or completely replace that business process. And then starting to work with the different parts of the organization, starting small and scaling it to help drive that adoption in a more efficient manner. And then of course, there's also all sorts of types of training and everything that we provide along that journey as well.Johnston: Is this used mostly by major companies or large companies or can it be used also within, let's say a mid-market company?Sax: We've got customers that have 20 engineers that utilize this, we have customers that have 5,000-plus engineers that utilize this tool. So the value and the impact that it's going to bring is going to be the same just going to scale with the size of the business, but it's still there. No matter the size of the organization, the impact and the results that it's going to bring are gonna be similar.Johnston: Ok. Well, thank you, Evan. I'm afraid that is all the time that we do have for today. So I'm gonna go ahead and wrap things up. If we didn't get to any of your questions today or you still have one, feel free to drop it in. We'll be sure to follow up as soon as we can. Evan, I want to thank you for giving us your time and your subject matter expertise today. Really appreciate it. I'd also like to thank you for sponsoring today's webinar. Once again, the recording will be available ConsumerGoods.com. And so finally, thank you to our attendees for giving us your time today. We hope you found it valuable.

Related Topics