Name: The Green Potential of Knowledge Graphs
Uploaded: 2025-11-06T17:02:53.520Z
Duration: 12 min
Description: The Green Potential of Knowledge Graphs

Transcript for "The Green Potential of Knowledge Graphs": Oh, yeah. Yeah. Okay. Just waiting for people joining. Mhmm. Okay. I think we can start now. So hi, everyone. Joaquin, you're the presenting this session today. So, Joaquin, over to you. Thank you. And, thanks everyone in listening in to this talk and my presentation titled, the green potential of knowledge graphs. Just a quick background of myself. I'm my name is Jorgen Nielsen. I'm a knowledge craft lead at Capgemini and also a client partner lead for Neo4j, across Europe. I'm based in Malmo in Sweden. I'm sitting right now about 50 meters away from the Neo4j headquarters, actually, in our Capgemini office. I have experience, from Neo4j, running projects both in Sweden and abroad, for public and private organizations across several industries. And I have my educational background in mathematics. In my role as client partner lead for Neo4j, I have seen a very strong interest for Neo4j for many use cases. One of which is the use of graph technology in sustainability. Both as a means to identify and reduce sources of emission, but also to make ESG reporting easier. During this summer, I co authored an article together with Neo4j and AWS on how knowledge graphs can help organizations gain a better understanding of their CO two emissions, which is crucial both from a legislative point, but also in order to take action and reduce your carbon footprints, you need to know where to start. What are the main drivers? And see down to component level, how much is that contributes to the overall CO2 emissions. In this article, we decided to focus on supply chains. CO2 emissions are a leading driver of climate change, and supply chains are a major, contributor. Every stage from raw material extraction to manufacturing, transportation, packaging, and distribution adds emissions, often in ways that are hard to measure and manage. Supply chains account for over 60% of global greenhouse gas emissions, and transportation alone contributes about 14. These emissions are not only harm the environment, but also create financial and reputational risk as regulations tighten and consumer expectations rise. Modern supply chains are vast and complex, spanning continents and involving thousands of suppliers and multiple transportation modes. Data is fragmented across ERP systems, logistic platforms, and sustainability tools, making it difficult to identify emissions and hot spots or evaluate greener alternatives. To address this, organizations need visibility, traceability, and actionable insight into their value chain. And this is where knowledge gaps like Neo4j comes in. Connecting desperate data sources into a unified relationship driven model that enables companies to understand and optimize supply chain for sustainability. And before going into real example, I that I did with a bill of material, let's just go through the main CO2 drivers, CO2 drivers in the global supply chains and how we can look. I decided to narrow down to three steps in the supply chain. First off, looking at manufacturing. So this stage involves parts, machines, processes, and energy sources, all contributing to emissions. By modeling relationships between the various components, processes, and emissions, and linking each part to its emission factor, energy usage, and supplier origin, we will end up with an enhanced traceability into questions such as which parts in this product come from high emission sources. What we see in this scheme is a knowledge graph representation of a manufacturing process where the we connect the product component identified by its IPN or internal part number. It's the starting point, of the manufacturing process. The manufacturing now then represents the specific processes used to produce the part. It's a central node because it connects to machines, suppliers, and emissions. The machine now links to manufacturing step through the use relationship, and this tells us which machine is involved in the process. This schema gives a gives us traceability and enables us to track emissions at the process level, not just at the product level. It allows scenario modeling, like what happens if we switch to a different machine or energy source. It supports compliance and reporting by linking emissions directly to processes and suppliers. In short, this graph turns a complex web of data into a clear connected model, enabling companies to identify high emission steps and optimize for sustainability. Another area is transportation, which is also a major CO2 driver in the global supply chain chains. Transportation is one of the largest contributors to supply chain CO2 emissions, and trucks, ships, planes, and trains have vastly different carbon footprints. Optimizing transportation for sustainability is complex. Routes, carriers, fuel types, and regulations create a web of interconnected decisions, which is made all simpler by using technologies like Neo4j. Here we can map routes, carriers, and transportation modes, each linking to each leg of the journey, to its carbon footprints, enabling scenario modeling like, what if we switch from air freight to rail. We can then use this for route optimization to find lower emission paths or comparing carriers to evaluate supplies based on their transport emissions. This next step we can go to is packaging. Packaging and distribution are often overlooked but contribute significant significantly to scope three emissions from packaging material production to warehouse energy use and last mile deliveries. And just to look at an example how we can do this. To demonstrate, I I took a bill material consisting of a product and broken down into its various component parts ranging from level zero to four. I broke out the headers, bill material, shortened boom here, part, site, and supplier from my Excel file, and turned them into nodes and user relationships, belongs to, has parent, has site, and supplied by, to create relationships between them. I store the c o two emission rates on the relationships between the different part nodes and aggregated up from level three to level two and so on up until reaching level zero. This way, we can easily see how the emission adds up as well as we as as we move up along the hierarchy. Here we have the bill of material node in blue connected to the highest level part node, level zero, with a belongs to relationship. From there on, we can traverse down the tree to level three with which has a with a has parent relationship going from level three to level two to level one and so on. In order to get a better visual understanding of how much each component contributes to the overall emission, I decided to use the Neo4j plug in NeoDash and its Sankey chart. So I think it gives a nice overview and flow of each component level to level. This, Sankey chart illustrates the c o two emissions across the components in in our bill of material. Each colored band represents a component, and the width of the band correlates with this emission impact. Components flow toward major assemblies. In this case, flex dry and thermosil, showing how emissions accumulate through the supply chain. Studying this chart, we can see, that the flex drive is a major convergence point with multiple components feeding into it. Its wide band suggests high cumulative emissions, making it a critical hotspot for optimizations. We can also see some more narrower bands which indicate minimal impact individually. But, collectively, if we were to have many of them, they still add up. To sum it up, implementing knowledge graphs in supply chain for sustainability purpose will enable us to identify and implement greener strategies, to find more effective routes for our shipping and reduce regulatory burdens by enhancing our traceability and explainability of our data. Reducing CO2 emissions in supply chain is not just about measuring, it's about acting. Knowledge graphs give organizations the ability to see connections that were previously hidden, but the real value comes from what happens next. Start by connecting ERP logistics and sustainability systems into a unified graph model. Use the graph to identify the highest emission processes, routes, and materials. Focus efforts where impact is the greatest. Simulate changes, switching transportation modes, adopting renewable energies, or using recyclable packaging, and quantify the benefits before investing. Share insights with suppliers and logistic partners to drive collective action. Leverage the graph for real time ESG reporting and to meet, scope three requirements without manual effort. It connects then the graph to include predictive analytics and AI driven recommendations for continuous improvement. The next step is clear. Move from visibility to action. Organizations that embrace knowledge graphs today will not only reduce emissions, but also build resilient future ready supply chains, opening the door to a greener future. That was all from my presentation. I'm just on time here. Thank you everyone for tuning in to listen. And if you have any questions, comments, or want to reach out, you will easily find me on LinkedIn. And, hope you enjoy the rest of the conference.