American academic, Thomas Davenport, said business processes are among the last remaining points of differentiation for firms that offer similar products and use comparable technologies. This concept holds especially true for the international trailer manufacturing scene.
Each of the world’s top OEMs take pride in their engineering capabilities to produce industry-leading, robust, efficient and productive road transport equipment and constantly inform the market about how they stand out from their direct competitors. While they all boast about their KTL painting — an anti-corrosion dip coating for heavy duty components — and aftermarket support networks to maximise vehicle uptime, It is clear that on an executive level these manufacturers are continuously fixated on selling their products by emphasising in-house innovation and their own tailored approach to customer service.
As the President’s Distinguished Professor of Information Technology at Babson College, co-founder of the International Institute for Analytics and a Fellow of the MIT Initiative for the Digital Economy, Davenport is an established authority on ‘analytics competitors’.
He outlined in his 2017 book, Competing on Analytics, that analytics competitors ‘wring every last drop of value’ from business processes to learn everything from how many items a client might buy in a lifetime to what triggers them to make a purchase or predict supply chain demand problems to improve inventory and order systems. These scenarios are countless.
Crunching numbers, taking names
Analytics competitors, according to Davenport, do all these things in a coordinated way “as part of an overarching strategy championed by top leadership and pushed down to decision makers at every level” and that employees hired for their expertise with numbers or trained to recognise their importance are “armed with the best evidence and the best quantitative tools”. He asserts that these individuals make the best decisions “big and small, every day, over and over and over”.
Davenport wanted to identify the characteristics that analytics competitors shared so he, along with two colleagues, studied 32 organisations that have made a commitment to quantitative, fact-based analysis. He found that 11 of those organisations were ‘full-bore analytics competitors’ i.e. top management had announced that analytics was key to the basis in which they compete — it’s in their DNA. Even though such a transformation requires significant investment in technology, Davenport claims that it is also equally important for executives to be committed and willing to change the way their employees think, work and are treated.
In one of numerous case studies, Davenport looked closer at US-based courier company, United Parcel Service (UPS), and said it embodied the evolution from targeted analytics user to comprehensive analytics competitor.
“The UPS Customer Intelligence Group, for example, is able to accurately predict customer defections by examining usage patterns and complaints,” Davenport said. “When the data point to a potential defector, a salesperson contacts that customer to review and resolve the problem, dramatically reducing the loss of accounts.”
Ultimately, for a company to embrace analytics that will drive change in everything from culture to process to behaviour, Davenport is adamant that it requires a principal advocate like a CEO, naming Jeff Bezos as an example.
Bezos, at the head of e-commerce firm, Amazon, is diversifying his portfolio with the development of cloud-based solutions that leverage Artificial Intelligence (AI) and machine learning. In February, Amazon Web Services announced Second Spectrum, the Official Optical Tracking and Analytics Provider of sports leagues such as the National Basketball Association (NBA) and the English Premier League, chose AWS as its preferred cloud, machine learning and AI provider. With innovation comes opportunity.
US-based transport and logistics company, XPO Logistics, developed and implemented its ‘XPO Smart’ workforce productivity tools in its less-than-truckload (LTL) network, last year, in North America to make dock operations more efficient. The proprietary analytics and optimisation tools compare real-time productivity rates with the number of active dock workers, using machine learning to predict how adjustments in labour levels affect productivity. The tools track motor moves (goods moved from dock to truck) against production targets and analyse productivity gaps to improve performance.
Meanwhile, materials handling specialist, Honeywell Intelligrated, unveiled an automatic robotic solution for unloading packages from trailers using AI to operate fully autonomously inside of a trailer, which according to Honeywell, significantly reduces the manual effort required to operate receiving docks for retail merchandise and parcel distribution centres.
“For distribution centre workers, unloading packages is labour-intensive, physically demanding and injury-prone work that is often subject to extreme temperatures,” said Matt Wicks, Vice President Product Development at Honeywell Intelligrated. “With our robotic unloader, we are using advanced machine learning to allow workers to remove themselves from the extreme environment and to oversee multiple unloading machines, increasing productivity and improving safety,” he said.
Honeywell’s robotic unloader drives into a trailer or container and uses machine vision to identify various package shapes and sizes as well as the optimal approach to unloading. A robotic arm with a series of small suction cups conforms to the package shape to gently extract it from the stack. A conveyor below the arm can serve as a sweeper for packages to move them out of the trailer.
“In real-world applications, we are unloading a rate of up to 1,500 cases per hour and helping companies maximise throughput safely and efficiently,” Wicks said. “We’re working with Carnegie Mellon University to deploy advanced machine learning to expand the robotic capabilities with improved 3D vision, perception, processing power and gripping.”
The Honeywell Intelligrated unloader is fully autonomous and is designed to work within existing fleets to eliminate the need for costly configurations or modifications to trailers or standard shipping containers. The unloader features patented gentle mechanisms to minimise package damage without impacting performance.
Last year, US-based commercial fleet management company, Ryder System, was recognised for providing cloud and mobile-based technologies that support a fully optimised supply chain with greater speed-to-market performance for food and beverage customers. The company’s innovative technology solutions include RyderShare, which is aimed at creating real-time visibility, collaboration, and exception handling to maximise efficiencies for customers. In addition, Ryder’s OpsBox is a warehouse analytics platform for labour management, providing shop floor visuals, daily metrics, workforce planning, and customer visibility dashboards. The result is labour productivity improvements exceeding 10 per cent into the double digits, as well as increased employee engagement and improved customer experience according to Ryder.
Extracting patterns that will persist
By establishing an analytics culture in a business — instilling a company-wide respect for measuring, testing, and evaluating quantitative evidence — the possibilities seem to be endless. Davenport’s work continues to influence executives and academics worldwide.
Dr Nigel Clay, Research Fellow at RMIT University, has lectured on the subject of machine learning, taking students ‘under the hood’ of algorithms and codes, and has also built models for fun and worked on them to solve business problems.
Clay said machine learning is not well defined as there is an emphasis on the learning however he characterises it as a blend of statistics and computer science, with a focus on extracting patterns that will persist.
“Information extrapolated from a machine learning model could, for example, help identify what factors might contribute to someone defaulting on a bank loan or why an employee might leave their job,” he said — adding that there are a lot of ‘canned’ approaches to machine learning. “There is an overlap between machine learning and AI but ultimately the pursuit is finding patterns that generalise.”
These patterns are particularly of interest for those embedded in supply chain optimisation. “In order to find the most efficient routes you need to test every possible route,” Clay said. “The more destinations you add, the more the routes increase. There are tricks to discount or spare the evaluation of certain data sets if you know the plausible outcomes are not in those selected areas. When dealing with the fuel efficiency of a fleet of longhaul trucks or courier vans, for example, effective route planning not only ensures timely delivery but also cost savings due to fuel consumption.”
Machine learning can also be applied to maintenance scheduling and predictive maintenance in a production-line environment according to Clay. “Knowing where you are on the failure pathway can help a business react to the situation as quickly and consistently as possible,” he said.
Similarly, in terms of condition monitoring, infrared thermography via a thermal camera is frequently used in factories for predictive-maintenance purposes. These cameras monitor the temperature of electrical and mechanical systems so that trained personnel, thermographers, to ensure onsite machinery is performing within safe and efficient parameters.
The thermal imaging camera typically works by interpreting radiated energy from an image and showing the temperatures as colours. The hotter the section of an image is, the colour will be a vivid red that transforms to white whereas the colder a section is, the darker the image will be (dark blue, verging into black). In this instance, machine learning and image processing could theoretically complement thermographic condition monitoring to automate the task. More complex actions could include the automation of follow-up predictive maintenance activities such as bearing replacement via robot. The possibilities really are endless.
In addition to the manufacturing and supply chain applications of machine learning, Clay points to finance as a major opportunity for businesses to refine their operations. “Machine learning can be directly applied to the leased residual value of vehicles,” he said. “Data sets including auction house data, total kilometres travelled and the age of the asset can be used to find trends in trailer resale value and perceived market value. These patterns can also inform product specialists on minimum guaranteed finance, bullet payments and other critical variables.”
An efficient, data-driven approach to vehicle resale, leasing and insurance would be a powerful resource/advantage for any trailer manufacturer or equipment leasing firm.
Chasing positive change
Those who do not record history are doomed to not learn from it, Davenport mused in Competing on Analytics. He said companies that have collected little information, or the wrong kind, will need to source a sufficient body of data to support reliable forecasting. He quoted a UPS Manager of Customer Data Analytics on the matter: “We’ve been collecting data for six or seven years, but it’s only become usable in the last two or three, because we needed time and experience to validate conclusions based on the data”.
Becoming an analytics competitor, applying the latest and emerging technologies, backed by reliable, useable data seems to be the logical conclusion for business executives to peak in their respective industries.