Previous Submission

Can Big Data Save The American Machinist?

Next Submission

Paperless Parts harnesses big data to give local machine shops a big boost through rapid quotes and access to an Amazon-like marketplace.

Massive growth in U.S. startups have driven the need for prototyping and smaller scale manufacturing.  With over 10% growth in 2017 alone, the need for greater access to both additive and stock removal manufacturing is equally great [1].  Fortunately, domestic manufacturing facilities already exist to meet the growing need, however, these small facilities often don’t understand how to quote new products delivered as CAD models, obfuscating a process that modern computing could easily simplify, turning off the younger startup culture [2].  Paperless Parts has built a machine learning engine to interrogate CAD models, identify efficiencies and provide real time quotes customized to the model, shop, and method of manufacturing [3].  Taking inputs of a parametric vector-based model and each supplier’s unique activity-based costing parameters, Paperless Parts’ algorithm utilizes machine learning to optimize the model to each shop’s equipment, past performance and manufacturing method to give the customer instant quotes for every type of production from CNC milling, to 3D Printing [4].

The marketplace Paperless Parts has created derives its true value from the instant and dynamic quotes it produces.  It allows the customer to optimize for every variable, from size, shape and quantity to shipping costs.  With 83% of all machine shops in the US employing 20 people or fewer [5], Paperless Parts’ ability to connect these underutilized, localized shops with customers often located on coastal startup hubs is valuable, but the savings to shops that often employ 1-2 people solely to quote potential jobs is a massive expense that Paperless Parts is able to solve through machine learning.

Fortunately, Paperless Parts value added is not tied directly to revenue.  Every quote produced will continue to add data, iteratively refining their algorithm, both increasing accuracy and driving down costs.  Additionally, every quote produced can be used by each shop as a benchmark to continue to increase efficiencies, and consequently, affect their costing parameters creating a virtuous cycle for both manufacturers and consumers.

Paperless Parts has short term goals of continuing to partner with more and more shops nationwide [6].  Each new shop, with unique assets and costing parameters refines the algorithm with each quote produced and adds granularity to the accuracy with which they can quote.  It has the added benefit of relieving quoting responsibility from the shop itself, freeing up labor and capital.  Additionally, they intend to add capability to factor economies of scale to account for repeat orders and just-in-time ordering to give the most accurate pricing and costing information to each party.

Longer term goals are to include options for more manufacturing processes, catching up with and staying at the forefront of the additive manufacturing sector.  As the industry moves past FDM and SLA/SLS, Paperless Parts will develop the model interrogation and optimization algorithm to further optimize each product produced.  Additionally, Paperless Parts will produce a tool that will enable a company to use their algorithm to determine whether or not to bring production of a product in-house, by allowing them to compare manufacturing costs against the cost to purchase and operate the machine themselves [7].  As more and more materials are available in the additive sector, nearly autonomous manufacturing may be possible and efficient for some companies and Paperless Parts can help them understand that.

There are a few ways Paperless Parts can improve their offering to both add and extract more value from the process, the first being a more refined method of enforcing and advising DFM (Design for Manufacturing) [8].  A guiding principle of more established product designers, it is often something missed by newer startups, but an algorithm that not only shows the customer which elements of their part contribute to higher cost or inhibit a specific type of manufacturing, but a tool that utilizes the large data set to extrapolate more efficient designs for the customer.

Additionally, the should build out their machine learning algorithm to tag each element of each part to refine the granularity of the data set and give them the ability to optimize products from best practice while preserving the integrity of intellectual property.

Paperless Parts should, upon reaching a large enough scale of orders, use their size to leverage favorable rates with freight carriers across the country.  Because they include shipping in their price quote, this could give them a significant competitive advantage that would flow directly to both manufacturers and customers, increasing both sales and ultimately profits.

Paperless Parts is well positioned to take advantage of an antiquated system of quoting and accessibility gap in US manufacturing, but can they include and retain enough manufacturers to provide a competitive offering over possibly cheaper options in the Asian market?  Can they scale in a way that enables them to continuously offer better and more refined tools while maintaining profit and positive cash flow?

 

[796 words]

 

[1] Source: [Rate of Startups Worldwide by Region], via Statista, accessed [11/2018].

[2] Jason Ray, Founder’s Spotlight: Jason Ray of Paperless Parts, interview by Emma Wright, CIC 2017.

[3] WTFFF: 3D Printing Tips and Tricks, “Finding a 3D Print Manufacturer,” podcast, https://3dstartpoint.com/finding-a-3d-print-manufacturer-with-jason-ray-of-paperless-parts/

[4] Ibid.

[5] Peter Zelinski, “Are U.S. Machine Shops Choosing to Stay Small for the Long Haul?,” Modern Machine Shop mmmsonline.com, 1 December 2017, accessed November 2018.

[6] WTFFF: 3D Printing Tips and Tricks, “Finding a 3D Print Manufacturer,” podcast, https://3dstartpoint.com/finding-a-3d-print-manufacturer-with-jason-ray-of-paperless-parts/

[7] Ibid.

[8] Ibid.

[Featured Image] https://www.3dhubs.com/knowledge-base/cnc-machining-manufacturing-technology-explained

3 thoughts on “Can Big Data Save The American Machinist?

  1. Very interesting article about utilizing excess capacity of machinists. One point I have regarding the scaling of this company is the variability in machining. Unlike commoditized industries, where the sharing model has been successful, the machining sector has large variances in the skill, capability, and speed of producing goods. Being able to communicate all these aspects to both the consumer and producer poses a significant challenge. If they are able to solve this problem, what is next? Mechanics, contractors, and other types of labor could be sourced and advertised on a marketplace forum.

  2. Very interesting article and company! I found it fascinating how dispersed the American machinist shop industry is and see forging connections with those shops as an enormous competitive advantage to be leveraged. I think that your suggestion of using machine learning to offer customers suggestions on improving designs to achieve cost efficiencies for manufacturing at scale to be incredibly interesting. I wonder how far machine learning is from being able to analyze CAD drawing and generate designs alteration ideas that stay true to the original need/intent while providing cost savings.

  3. If I understand it correctly, Paperless Parts works as a market place for production capacity in machine shops. Thus, Paperless Parts becomes customer-facing and the machine shop is reduced to the mere production unit. We have seen this business model to be successful in other industries, e.g. in the printing industry. However, there are distinct differences in terms of the customer-supplier relationship between the machine job industry and other industries: trust. In many cases customer relationships have been built for decades. The trust customers have in their machine shop seems to be more important than the x% cost reduction which might be generated with Paperless Parts and sophisticated machine learning tools. In contrast to prints, machined parts will be built into the final product for which the customer is reliable. In fact, very often the lower-value parts are outsourced to small machine shops, so the potential cost reduction might even be negligible relative to the value of the final product.

Leave a comment