Industry
Manufacturing
TCS Neural Manufacturing: Driving Agility & a Competitive Edge
Highlights
- Scope 3 emissions, which have the widest range, are difficult to measure because of inherent complexity.
- The conventional bottom-up data analysis method is expensive, time consuming, and not scalable.
- Sustainability requires an end-to-end view with real time, real world, consistent, and accurate data for decision making across the value chain.
- Integrating such data points can help companies in their net zero journey.
On this page
The data challenge
Delivering net zero goals and embedding circularity in manufacturing are huge data challenges.
Measuring, tracking, and reporting transparently on the carbon footprints of trillions of items sold to billions of people, produced across millions of locations from tens of thousands of components is not easy. Currently, in many manufacturing organizations, the use phase or end- of- life data is assumption-based. Generic data sets for carbon dioxide emissions don’t accurately represent the supply chain or allow for variations in product specifications. Combined scope 3 emissions by derivatives or parts are restricted and limited, as at that level, the data is mostly aggregated. Scope 1, 2, and 3 impacts are reported with little or no mention to other environmental impact.
An automotive story
Understanding the data complexities and other factors that hamper intelligent decision making is the first step.
Let’s take an example of an automaker that manufactures 450,000 units in a fiscal. This translates to a minimum 2,000 unique derivatives or parts, about 700 materials, (classified as grades of steel, aluminium, various plastics, rubber, glass, and textile) and over 5,000 suppliers requiring multiple modes of transportation and deliveries, both ad hoc and planned. The manufacturer has multiple manufacturing sites across continents with a similar number of varied production lines.
Besides these complexities, there is a need to track changes in car composition, battery traceability, recycled content, employee, and community welfare, and greenhouse gas (GHG) emissions too. Data, in these cases, is fragmented across multiple functions giving a point- of- time view at best for a limited area. Multiple sites use multiple systems with each managed independently and with a focus on compliance. There is a distinctive lack of intelligence to drive improvement and efficiencies.
The visibility conundrum
The conventional approach to any solution is a form of bottom-up data analysis where we start with a single, defined product that is the subject of the study.
The approach, where materials and processes used to manufacture a product are systematically mapped back through the value chain to quantify their environmental impact, is nonetheless a complex task. It requires specialist knowledge and sophisticated software tools. It is usually time-consuming, expensive, and not scalable across an organization, which deals with multiple products.
Every department within a manufacturing unit needs visibility into the material – whether it be to source, manufacture, transport, recycle, or simply report on it. With growing regulatory requirements, commitment to Science Based Targets initiative (SBTi) goals, and governments looking to control sustainability footprints, there is a need to bring a cohesive view of enterprise data.
This integrated view will allow for scalable and autonomous reporting, proactive monitoring, and reduction of emissions. It will also conduct a what-if analysis into the future with simulations to measure the impact on production balance, stock, procurement, GHG emissions, and carbon intensity. Without that cohesive view, all these advantages are lost to the decision makers.
Data considerations
There is a need to look at sustainability with an end-to-end view.
And for that, manufacturers need real-time, real-world, consistent, and accurate data to inform decision- making across the value chain. The ability to integrate billions of datapoints from complex value chains to enable short and long-term decision making on delivering net zero goals will transform the very essence of any business model. The key elements to keep in mind are:
- Mining existing data: All data that is needed to get deep insights into the lifelong impact of cars on the environment can be found in existing systems. Purchase data, bills of material, energy consumption, and production volumes can be captured directly from management systems and there is no need to generate expensive new datasets.
- Speed and scale: All automotive companies have been using the life cycle assessment (LCA) technique for a long time. The process, however, is time- consuming, and output is limited in terms of the insights it gives to those teams actively trying to find ways to achieve net zero. So, the objective is to deliver LCA results for multiple products simultaneously, repeatably, and at a level of detail that will enable robust decision-making.
- Insights based on real data: Unlike the traditional, expensive, generic, and slow bottom-up LCA methodology, the industry should look at activity-based foot printing, turning the tide on traditional LCA.
- Strategic rather than tactical: Businesses need data around future, and circular business models with impact on biodiversity and climate. The ability to utilize data and power it by artificial intelligence (AI), machine learning (ML), digital twin, and industry-tuned models can help perform multi-dimensional simulations that will allow decision makers to test future options across their value chains based on thousands of variables before investment decisions are taken.
At a time when trillions of items, billions of consumers, and a large number of suppliers are making relentless efforts to become net zero across marketplaces and value chains, it's vital to use new-age technologies to not just gain a rear-view-mirror perspective of our historic performance, but also to identify and optimize our pathway into the future.
