RPA in manufacturing has been saved
RPA in manufacturing
Using cognitive analytics to prevent parts proliferation across industries
Parts proliferation is a persistent problem that can cost manufacturers hundreds of millions of dollars every year. But advances in AI, cognitive analytics, and robotic process automation (RPA) are making it easier, helping manufacturers manage parts complexity with tools and processes that can quickly deliver strong ROI.
- An evergreen proliferation problem
- Design Advisor
- RPA in manufacturing parts management
- Get in touch
An evergreen proliferation problem
Parts proliferation significantly increases costs and undermines efficiency. The impact can be felt throughout the supply chain in the form of poor purchasing leverage, poor reuse, and increased levels of inventory.
In addition, there are numerous downstream impacts on other business areas, including:
- Demand planning
- Manufacturing planning
- Supplier management
- Customer support
Proliferation also adds complexity that makes it harder to understand the true cost of design changes. These inefficiencies across hundreds of thousands of redundant parts create the potential for massive cost savings.
Traditional human-based approaches to parts rationalization often leave a lot of money on the table. Even worse, they must be repeated periodically or even continuously, since they can be connected to traditional design processes and systems that continue to foster parts proliferation.
A better solution: Cognitive analytics-and RPA-driven part rationalization with advanced search
Our twofold proliferation solution: DesignSource™
To solve the parts proliferation problem once and for all, manufacturers need to rationalize their parts inventory using a system that can automatically extract or digitize parts data from a wide range of structured and unstructured sources. Once complete, this allows an organization to enable advanced cognitive searches beyond those available in PLM solutions to make the product and supply chain data instantly accessible to engineering.
Deloitte’s DesignSource™ is a rigorous analytical tool that automatically analyzes parts data across functional silos to give design and procurement teams the data-driven insights they need to optimize direct material spend and cost of goods. Key capabilities include:
- Collecting product, design, purchasing, supply chain, and operations data from a variety of sources
- Merging BOM data with other enterprise, supplier, digital supply chain, and third-party benchmark data
- Cleansing and enriching part attribute data from manufacturing part numbers and unstructured sources, such as drawings and data sheets, using machine learning–based automated tools
- Identifying clusters of similarly functioning parts for consolidation, price harmonization, and optimized sourcing
- Identifying preferred parts that exhibit desired design standards and technology, allowing for classification of BOMs for reuse
- Deploying prebuilt should-cost models across thousands of parts and categories
Our twofold proliferation solution: Design Advisor
As part of the rationalization process, advanced cognitive analytics are used to identify parts that are near-duplicates, enabling a level of streamlining and cost savings that far exceeds what traditional human-based rationalization efforts typically produce.
A preconfigured Deloitte solution, Design Advisor, is an RPA-driven machine learning and analytical solution that proactively evaluates and recommends parts and designs in real time. Design Advisor cognitively compares the requirements of new designs to existing designs, then provides suggestions for reusing existing designs, using known procurement cost and technical and supplier information to support those suggestions.
RPA in manufacturing for advanced parts search also has the intelligence to analyze and extract information from unstructured data sources like technical drawings and documents to make reuse recommendations.
These real-time capabilities make a designer’s job much easier and reduce the temptation to design or specify a new part when an existing part would work just as well. To reinforce this, Design Advisor incorporates prompts embedded in new design processes using PLM change management workflows to “catch” engineers during part creation for a justification or approval.
RPA in manufacturing parts management
RPA in manufacturing parts management can help organizations:
- Achieve cost visibility across similar parts in hours and realize savings within a few months
- Fund ERP and PLM technology programs that touch the same product and supply chain data through savings generated by DesignSource™
- Rationalize part numbers and suppliers under management
- Drive volume consolidation through reuse from preferred suppliers at negotiated prices to minimize new product introduction costs
- Create actionable road maps to capture benefits over short, medium, and long term through negotiations and engineering changes
- Save engineering time identifying similar designs and streamline resources for value-added activities
- Reduce parts complexity management throughout the organization by enforcing usage of preferred parts
- Reduce material spend as procurement can source higher volumes on fewer designs
Typical cost savings range from 4% to 12% of procurement spend, which for some companies can translate into tens of millions of dollars annually.
For example, a life sciences company used RPA-driven parts rationalization to identify 5% (about $19 million) in direct material savings that could be achieved through short-term price harmonization, consolidated purchasing, and improved design standards.
Similarly, a major industrial products manufacturer used RPA-driven parts rationalization to achieve >$30M in savings across 18 commodities and $530 million in spend for off-the-shelf and semi-engineered parts. The company digitized 600,000 unique parts with part attributes from roughly 100,000 drawings and data sheets. By clustering parts, 30% of parts were identified as duplicates for elimination in the near term. Also, recommending preferred parts for search and reuse had the potential to reduce the company’s overall parts portfolio by 47% over the long term.