How much value does Machine Learning deliver to traditional industries? Although vital to the operation of almost all technology companies and hot start-ups, Machine Learning seems to only be adjunct to the core businesses of traditional manufacturing and operations. Having worked in multibillion dollar oil and gas project development, construction, and operation, I see Machine Learning having immense untapped potential in the sector both during the product development phase and process improvement, including after sales and maintenance. To this extent, Caterpillar is perhaps one of the few companies delivering bottom line savings using Machine Learning in a consistent manner.<\/p>\n
Through its Asset Intelligence Platform, Caterpillar uses data analytics to identify seemingly unrelated correlations to deliver optimum performance. For example, fuel meter readings are used to gauge efficiency and resulted in comparative savings when operating all refrigeration generators on low power rather than maxing out a few, leading to aggregate yearly savings of $650,000 [1]. Another example in their marine operations is hull cleaning optimization. Through adaptive learning of the efficacy of certain cleaning maneuvers to remove barnacles and seaweed, Caterpillar has been able to firstly establish an optimum cleaning motion and secondly trace a relationship between cleaning costs and performance improvements. The end result is more frequent maintenance (reduced from once every 2 years to once every 6.2 months) that improved performance and lead to $400,000 yearly savings per ship [2]. In addition to the aforementioned improvements, Caterpillar also has $9 to $18 billion in easy to capture revenue by commercializing real-time customer data, mediating inter-user communication, and standardizing customer experiences [3].<\/p>\n
Caterpillar\u2019s management has set its dealers a one year deadline to produce a 3-year action plan to capture lost sales from distribution. This short-term program incentivizes dealers to install data collection systems and provide service sales whilst simultaneously capturing revenue lines by scheduling preventative maintenance and efficient fleet management through Artificial Intelligence [4]. Moreover, Caterpillar estimates that it should be able to collect data from the 3.5+ million pieces of equipment currently in the field by integrating them in the smart digital platform vis a vis their Across-the-Table initiative[5]. In addition to short term initiatives, Caterpillar has put in place a strategy to integrate Machine Learning in the medium to long term.<\/p>\n
Caterpillar has set three strategic targets to help sustain its value creation initiatives in Machine Learning over the next 10 years. Firstly, Caterpillar is funding resources in Silicon Valley via innovation labs and venture capital seeding to extract value from emerging technologies within the next 10 years [6]. Secondly, Caterpillar is seeking to enter partnerships with large firms to identify common synergies following the footsteps of Komatsu and GE [7]. Finally, Caterpillar has set up a new Artificial Intelligence (AI) unit that will be lead by Doug Hoerr, the ex-vice president of the strategic services division. Hoerr\u2019s responsibilities include strategic investments, mergers and acquisitions, and strategic planning thereby setting the pace for sustainable revenue generation from Machine Learning and AI integration [8].<\/p>\n
Recommendations for Caterpillar in the short-term span two domains, breadth and depth. Firstly, Caterpillar must seek to apply Machine Learning to all its core business lines through its new AI business unit by integrating operating agents in separate business lines and centralizing findings for the purposes of cross-application. Moreover, Caterpillar should seek to gain breadth by courting its current consumers to install data collection devices, even if at below cost, such that it is able to increase its current data pool and become more effective in providing tailored solutions and finding correlations that lead to significant savings. Caterpillar can then commercialize on these savings through business proposals on a customized basis.<\/p>\n
In the medium term, Caterpillar must seek to grow and monetize its AI business unit. Growth should be achieved by investments along the value chain of innovation, from research labs to direct funding of start-ups, partnerships, and acquisition. Monetization could be achieved through one of two ways. Foremost, Caterpillar should commercialize its data collection mechanisms through either spinning-of the AI business unit as a consulting subsidiary of Caterpillar or a consulting branch wherein it can further realize returns by providing services and selling products and services on a business to business level.<\/p>\n
Finally, two questions remain as to the sustainability of Machine Learning in a traditional industry: To what extent will Machine Learning be viable given the diminishing returns from capital investment? And how can Caterpillar maintain its first mover advantage as a source of sustainable value creation going forward?<\/p>\n
(Word Count: 747)<\/p>\n
[1] Marr, B. (2018). IoT And Big Data At Caterpillar: How Predictive Maintenance Saves Millions Of Dollars. [online] Forbes. Available at: https:\/\/www.forbes.com\/sites\/bernardmarr\/2017\/02\/07\/iot-and-big-data-at-caterpillar-how-predictive-maintenance-saves-millions-of-dollars\/#368fedfd7240 [Accessed 7 Nov. 2018].
\n[2] Bekker, A. (2018). Big data in manufacturing: 12 real-life use cases. [online] Scnsoft.com. Available at: https:\/\/www.scnsoft.com\/blog\/big-data-in-manufacturing-use-cases [Accessed 5 Nov. 2018].
\n[3] Bekker, A. (2018). Big data in manufacturing: 12 real-life use cases. [online] Scnsoft.com. Available at: https:\/\/www.scnsoft.com\/blog\/big-data-in-manufacturing-use-cases [Accessed 5 Nov. 2018].
\n[4] Casey, S. (2015). Bloomberg – Caterpillar Focuses on Big Data with Creation of Unit. [online] Bloomberg. Available at: https:\/\/www.bloomberg.com\/news\/articles\/2015-04-10\/caterpillar-focuses-on-big-data-with-creation-of-unit [Accessed 5 Nov. 2018].
\n[5] Yahoo Finance. (2015). Caterpillar Joins the Big Data Bandwagon, Creates Unit – Analyst Blog. [online] Available at: https:\/\/finance.yahoo.com\/news\/caterpillar-joins-big-data-bandwagon-204508643.html [Accessed 6 Nov. 2018].
\n[6] Caterpillar.com. (2018). Caterpillar | Caterpillar Ventures – About Us. [online] Available at: https:\/\/www.caterpillar.com\/en\/company\/innovation\/caterpillar-ventures\/about-us.html.html [Accessed 5 Nov. 2018].
\n[7] Nikkei Asian Review. (2015). Komatsu, GE to offer big-data analysis for mining projects. [online] Available at: https:\/\/asia.nikkei.com\/Business\/Komatsu-GE-to-offer-big-data-analysis-for-mining-projects [Accessed 5 Nov. 2018].
\n[8] Caterpillar. (2015). Caterpillar | Caterpillar Announces New Analytics & Innovation Division. [online] Caterpillar.com. Available at: https:\/\/www.caterpillar.com\/en\/news\/corporate-press-releases\/h\/caterpillar-announces-new-analytics-and-innovation-division.html [Accessed 6 Nov. 2018].<\/p>\n
<\/p>\n","protected":false},"excerpt":{"rendered":"
How much value does Machine Learning deliver to traditional industries? Although vital to the operation of almost all technology companies and hot start-ups, Machine Learning seems to only be adjunct to the core businesses of traditional manufacturing and operations. Having […]<\/p>\n","protected":false},"author":11020,"featured_media":29269,"comment_status":"open","ping_status":"closed","template":"","categories":[4365,346,2373,344],"class_list":["post-29229","hck-submission","type-hck-submission","status-publish","has-post-thumbnail","hentry","category-artifical-intelligence","category-machine-learning","category-process-improvement","category-product-development","hck-taxonomy-organization-caterpillar","hck-taxonomy-industry-construction","hck-taxonomy-country-united-states"],"connected_submission_link":"https:\/\/d3.harvard.edu\/platform-rctom\/assignment\/rc-tom-challenge-2018\/","yoast_head":"\n