The fundamental difference between IoT analytics and traditional analytics lies in the former’s ability to perform analytics and generate insights at different levels of the IoT network hierarchy, while the latter focuses on doing this at a central location. As a result, IoT analytics presents a unique ability to generate insights at the point of consumption and enables a truly distributed decision-making approach.
IoT analytics utilizes the ability of the sensory and supporting hardware, actuating devices/controllers and hierarchical intelligence to analyze local data, draw locally relevant insights/predictions and implement local decisions.
Take, for example, a vehicle outfitted with sensors that can collect and interpret data on the wear and tear of critical parts (e.g., tires) and share these insights with the car owner. At a higher level, this data can be aggregated across vehicles to predict when tires need to be changed, reducing the possibility of accidents and lowering the cost of individual vehicle ownership. On a regional level, the analysis of critical parts data can provide insights about tire demand to dealers and manufacturers to help supply chain partners work more proactively to anticipate and meet impending demand.
Combining IoT data with analytics provides manufacturers with a 360-degree view of operations and a significant monetization opportunity. This extends beyond traditional end customers to the larger ecosystem of partners and stakeholders.
The monetization potential extends to a much larger ecosystem of stakeholders (see Figure 1). This comprises an “inner sphere” consisting of players directly connected with the manufacturers (e.g., for the auto industry they include suppliers, telecom providers, dealers and end customers) and an “extended sphere” consisting of players not directly connected but which support the manufacturer’s product ecosystem (for the auto industry these include insurance providers, finance companies, retailers, government agencies, etc.).
Analytics, in general, can help manufacturers drive financial improvements through improved performance metrics. For instance, predictive analytics can optimize costs for manufacturers. Cost savings of 2% to 4% from a 50% penetration of IoT in manufacturing can deliver $500 billion in cost savings (assuming the global cost base of manufacturing at $25 trillion). Rio Tinto, a mining major, saves over $300 million by deploying “autonomous mining.” Similarly customers of contract manufacturer Flextronics are now positioned through IoT analytics and improved market visibility to quickly react to irregularities in supply chain components by performing real-time data correlation.
According to IDC, 55% of discrete manufacturers have undertaken IoT initiatives (i.e., research, pilots and adoption). The rapid rise of IoT in manufacturing makes it most attractive for the application of IoT analytics. Manufacturing is expected to account for more than one-fourth of the total IoT market (see Figure 2) with the oil and gas sub-segment of manufacturing leading IoT adoption along with the energy sector.
To succeed with IoT analytics, manufacturing companies must focus on the following critical dimensions:
Prioritizing customer requirements: Customer demand for high levels of service, experience and personalized products is pressuring manufacturers to quickly overhaul their business and production processes. For instance, Harley-Davidson reconfigured and equipped its facility in York, PA, with sensors and location awareness to reduce the time it takes to produce customized motorbikes from a 21-day cycle to six hours.
Getting the data/information right: Doing this requires manufacturers to first understand the kind of insights required by ecosystem stakeholders. An essential part of this is to clearly map all data elements in the network universe with the insights that are of potential use by stakeholders.
Advancing the IoT analytics value proposition: Embedding data/information/insights is an integral part of the systems and processes within the extended IoT ecosystem. Closing the loop on insights through the establishment of decision-making logic based on insights drawn from advanced analytics is hypercritical.
Realizing the value of insights: To derive value from proliferating IoT data, manufacturers must build insight models that are applicable at each level of the value chain. We view this as follows: informative » reactive » predictive » prescriptive.
To learn more, read our white paper “How Manufacturers Can Unlock Business Value via IoT Analytics” or visit the manufacturing and IoT sections of our website. For deeper insights, read our related Perspectives articles on IoT monetization and design-thinking.