Machines Learning: Caterpillar Inc.’s Metamorphosis into Big Data

Introduction: Today, there are 560,000 “connected” Caterpillar machines – from back-office forklifts to high-powered 50-ton excavators – that have been outfitted with connectivity solutions designed to cut costs and drive productivity for customers. [1] Caterpillar, the leader in machinery innovation for the last century, is using machine learning to harness an “overwhelming amount of data that [it] can now receive, process and feed back to [its] customers, dealers and factories in a way that [its] never been able to before”. [2] The world’s largest construction equipment manufacturer has begun its metamorphosis into a big data company.

Opportunity: The volatility of production inputs (raw materials, labor costs, regulatory shifts, customer preferences, etc.) in the global manufacturing environment is a significant risk factor for most businesses. Stabilizing input costs and driving productivity improvements are perpetual strategies for industrial organizations to win. Today’s confluence of digital technology and machine learning offers manufacturers and end-users of capital equipment an unprecedented opportunity to improve the performance of their assets. Predictive maintenance, a digital asset intelligence strategy that helps maximize utilization, minimize unscheduled equipment downtimes and predict failures, will be a key lever to lower operating costs. Manufacturers across the world are investing significant amounts of capital in predictive maintenance technologies to help transform their approach to maintenance, repair and overhaul (“MRO”). [3] Caterpillar’s growing portfolio of “Smart Iron” products and services position the company well as a value-added partner and leader in the evolving MRO market. The MRO market is expected to grow to $4 billion by 2022 in North America and Europe alone, presenting an attractive opportunity for Caterpillar to continue its metamorphosis by providing customers with savings that result from big data. [4] Additionally, end-users are increasingly seeking upstream partners with an ability to serve as a single point of contact to fulfill all of their MRO requirements, a trend that Caterpillar can leverage through its scale as a one-stop-shop for OEM and aftermarket services. [5] While many other equipment manufacturers are still in the design or early adoption phases of their connected technology initiatives, Caterpillar is well-positioned to transform the maintenance strategies of its customers from reactive to predictive and preventive.

Source: http://s7d2.scene7.com/is/image/Caterpillar/CM20170719-40349-06707

Management Actions: At Caterpillar, CEO Jim Umpleby emphasizes that priorities are shifting towards “what goes around the iron”. [6] Previously, only a portion of the equipment manufactured by Caterpillar was equipped with embedded connectivity. Now, every piece of machinery that leaves the factory is IoT-enabled. [1] In both the short- and medium-terms, Caterpillar must continue to refine its set of software and data analytics tools, as well as application user interfaces, to help customers understand how data and failure prediction can transform their MRO strategies. Caterpillar is also working to expand the breadth of its technology products and services to match customer needs. The company is creating a tiered offering of applications, ranging from “very basic on-machine GPS technology to fleet management and participation in a customer’s operations”. [7] To help manage its strategic shift into technology, Caterpillar’s management has formed a network of partners to provide clients with various options to process, analyze and store data. In 2015, Caterpillar took a minority investment in Uptake, a provider of big data analytics services, to jointly develop an end-to-end predictive diagnostic solution designed to help customers track and optimize fuel efficiency, uptimes, idle times and more. [8] Conversely, the company also appears to be building data analytics capabilities in-house. Its “Cat Asset Intelligence” service leverages an internal team to provide predictive analytics and advisory services for marine clients. Cat Asset Intelligence offers the technical flexibility to integrate with existing data sources and combine new sensors, lowering the costs for potential customers of switching to Caterpillar. [9]

Other Opportunities: Caterpillar management has done well staying ahead of the curve in capitalizing on the machine learning megatrend. However, there are still steps that can be taken to optimize its approach. Most notably, Caterpillar’s pricing strategy for its technology products and services appears nascent. Key data points like tracking hours, location and error codes, among others, are offered free of charge when purchasing a piece of equipment, and comments from Caterpillar management have indicated a lack of understanding around customer willingness to pay [10]. Caterpillar should perform tests (e.g., A/B) to better understand willingness to pay, which will inform a more sophisticated pricing strategy to capture growth. Additionally, the company should consider separating its new service offering from its traditional equipment manufacturing business.

Closing Thoughts: Caterpillar’s biggest challenge will be processing, analyzing and visualizing the sheer volume of data generated by its machines. This is not the company’s core competency. Should Caterpillar focus on building out its in-house capabilities like Cat Asset Intelligence, or should it outsource data analytics to industry experts like Uptake? Which approach is better long-term for cultivating the explosive growth in machine learning?

(800 words)

Sources:

Previous:

Google and Gmail: Machine Learning for Defense, Not Disruption

Next:

“Just let the Machines do the Work”: Marriot’s Opportunity to Build on Starwood’s Digital Expertise

4 thoughts on “Machines Learning: Caterpillar Inc.’s Metamorphosis into Big Data

  1. I believe that firms like Caterpillar, who don’t have Data Analytics as a core competency currently, should leverage a hybrid model where they outsource the development of the analytics and insource the strategic understanding of applications for analytics. This way they have the functional knowledge supplied by a trusted partner coupled with the industry knowledge within their organization. “The shortage of analysts — particularly those capable of developing and leading world-class teams that can enable a company to create a competitive advantage from its data and analytics — is driving organizations to consider outsourcing their analytics activities. However, choosing analytics providers and structuring effective working relationships that deliver value require managers to have a clear understanding of what they’re looking for and the potential risks involved (Fogarty & Bell) [1].”
    With Caterpillar’s minority investment in Uptake, it appears they are well positioned to follow this approach and use a partner with analytics as a core competency to effectively build their analytics capabilities. Per the commentary from Fogarty and Bell, I believe Cat Asset Intelligence and Uptake can augment each other as the company progresses their analytical offerings.

    Sources:
    [1] Fogarty, David, and Peter C. Bell. “Should You Outsource Analytics?” MIT Sloan Management Review, MIT Sloan Management Review, sloanreview.mit.edu/article/should-you-outsource-analytics-2/.

  2. Very interesting piece J Harvard! You pose an interesting question. Caterpillar is clearly a large company with an extensive fleet of equipment and machinery. As you mentioned, there is a lot of potential value add that Caterpillar can offer its customers through IoT and machine learning innovation. In order to differentiate itself, I think Caterpillar should develop its own in-house R&D and analytics expertise through by acquiring a company like Uptake.

  3. Thanks for the great read JHarvard! Predictive maintenance appears to be an exploding field for machine learning applications as similar technologies / algorithms are used in the aerospace industry, by firms such as GE, to ensure proactive engine maintenance.[1] In terms of your question on outsourcing vs. insourcing, I suggest Caterpillar review aerospace firms’ approaches as peer benchmarks to guide their decision. Caterpillar’s investment in Uptake struck me as an interesting middle ground that I am surprised others have not emulated. Given that big data analytics is not most firms’ core competency and data security is a massive risk, do you think more firms would be better served pursuing an acquisitive strategy rather than building in-house or outsourcing to an independent third-party? Lastly, I question how this market will be impacted by autonomous vehicles. Are we headed toward worker-less job sites? If so, will Caterpillar’s portfolio of IoT-enabled products become a competitive differentiator?

    [1] https://www.mro-network.com/maintenance-repair-overhaul/ge-aviation-steps-its-predictive-maintenance-efforts

  4. Thanks for the great post!
    I’m curious how they actually implemented this new service. I worked on several consulting projects with Bobcat, and for the most of the cases always the distributors were the biggest issue. (and they’ve also started the AI initiatives this year : https://www.constructionequipment.com/doosan-bobcat-looks-ai-equipment)
    One issue makes me concern in this case is that the dealers, or the company-owned-store employees be highly likely lack of the technical knowledge to explain the new features. In case of the dealers, I think CAT should find a way how to motivate the dealers to sell the new features, as well as to manage the end-price, or dealer’s margin. Also, I think it’ll cost a lot to set up a maintenence scheme – internal organization, policy etc, because as you pointed out, big data is not a field that CAT has been doing well.

Leave a comment