Executive Briefings

3 Strategies for Wrangling IoT's Big Data

Blaze a path through the maze of IoT data overload.

Among the weighty decisions leaders face about Internet of Things, much attention seems focused on choosing the right hardware and software, and where to install them.

“The equipment choices are important, absolutely,” says Charlie Covert, Vice President of Customer Solutions at UPS. “But IoT’s true value will be found in the data those devices carry. But having a clear vision can not only reduce IoT infrastructure expense, it can lead to data insights that can help to differentiate you in the market.”

Covert shares an example of what UPS Chief Information Officer, Juan Perez, calls the company’s “Innovation Cycle.” In that cycle, the technology built to support operational improvements also helps to create new UPS® services and capabilities.

“Analysis of the rich data captured by connected devices throughout our organization led to UPS Package Flow Technology, which evaluates and optimizes every step in a delivery cycle – that’s from the way it’s routed and tracked to the way it is loaded, and then delivered,” Covert says. “The visibility provided by Package Flow Technology then led to breakthrough customer solutions, like Package Intercepts, the UPS Access Point Network and UPS MyChoice service."

UPS’s Jack Levis adds the groundbreaking ORION route optimization system to the list of innovations enabled by Package Flow Technology. Levis, Senior Director of UPS Process Management, has helped to pioneer the system that uses advanced mathematics to crunch millions of data points per second, and guides UPS drivers on the most efficient daily route.

Says Levis, “I think the first step for everybody is, get your data in order. Understand the data you use to run your business, and gather all the data you can and start analyzing it. Whether you're a large company or a small company, leveraging the data you already have into meaningful insights puts you that much further ahead of the game.”

Prepare for the Data Avalanche  

For some perspective on the volume of data IoT will generate, an MIT Sloan report notes that sensors embedded in GE machines alone can collect “50 million data variables from 10 million sensors.” And those numbers are surely set to skyrocket. Global projections for connected devices in use by 2020 range from between 25 billion to almost 200 billion. 

Says Covert, “When you look at the amount of information that will be collected from various IoT technologies, it can definitely be overwhelming. But now is the time to take that first step,” he urges. “Start talking about the data you need, and set out a phased approach for how you’re going to get there.”

Levis suggests that companies address fundamental needs as soon as possible. “It would be unusual for midsize companies to already have on staff all the expertise they will need to manage and make sense of their data. Outside expertise can be invaluable for basic steps, such as auditing the data sources you have, and how to access and store the data,” he says. “The right questions to ask at this point are, ‘Do we need rackspace?’ ‘What analytics tools will we need?’ and equally important, ‘What staffing and expertise will we need to maintain an ongoing plan for data gathering, analysis and feedback?’”

Start with the Problem, and Work Backwards

The late Steve Jobs, master of turning the seemingly impossible into multi-billion dollar products, once said, “You have to work hard to get your thinking clean to make it simple.” 
 
Not to imply that it will be simple to find data that will propel your company into the next century, but the sentiment is appropriate to IoT, because of the shift in mindset that will be required. For example, leaders facing big projects are conditioned to gather reams of facts in order to determine a course of action. Instead, what if you took a different path? What if you detailed the things you want to accomplish and solve, and then worked backwards to identify where the data you need resides – or where it will reside once you get access.

Not only can working backwards from help you narrow down the parts of your operation to connect, you’ll be ready to use the data that eventually comes streaming back at you – those bytes and bytes and bytes of data.
 
“Starting with the problem or opportunity keeps you from solving a lot of little problems that really have no impact,” explains Jack Levis, Senior Director of UPS Process Management, and a pioneer of the ORION optimization system.

Evolve Your Analytics Capabilities over Time  

The ultimate success of Internet of Things initiatives, like any enterprise effort, relies on certain mainstays. For one, the effort will need long-term corporate commitment and funding to maintain momentum. Second, it must be backed up with a documented architecture and implementation plan, to make sure it can be carried out seamlessly regardless of personnel changes. 

Levis suggests adopting a progressively sophisticated approach to their analytics plan, which has become an industry best practice. “The first step, usually thought of as descriptive analytics, reports what’s already happened, where am I today?”

The next phase applies more complex analysis and algorithms that can make the insights more predictive in nature, or Levis’ words, “Based on what’s happened in the past, what will happen tomorrow?.”   

In the next phase of this long journey, Levis explains that the analysis becomes prescriptive, which may suggest preventive actions, determining the optimal locations for new facilities, or for determining inventory levels that maximize order fulfillment rates, while reducing waste.

Sometimes, the suggested actions may seem counterintuitive. In the case of UPS’s ORION, the program uses accumulated data to determine the most efficient driver route, and gets steadily smarter as it learns. At first, many experienced drivers balked at ORION’s suggestions, for example, to pass three stops and return to them later. In the end, drivers learned to trust ORION, and the proof is in the 100 million fewer miles driven last year with the same number of stops.

“People tend to think about what’s best to do right now, and then the next step, and so on. But analytics tools like ORION think of an entire network. Any individual decision may seem wrong, but it’s best for the whole day.”

Talk. Think. Transform. (Repeat)

The first step in any major initiative is often the most difficult. But now is the time to start. 

Talk – Get the conversation underway. If you’ve already started, give it a jump-start by bringing an extended group of stakeholders to the table. Digitally connecting a supply chain, from upstream to downstream, affects literally everyone in your ecosystem. Leverage their knowledge, and let them leverage your knowledge for mutual benefit.
 
Think – Once you’ve brainstormed with your stakeholders, assign task forces and deliverables. Build a business case that secures executive and Board-level advocacy. Draw on outside partners for specialized expertise. Create the architecture; identify milestones and checkpoints.

Transform – As might be expected, this phase is not a “one and done.”  Most companies will take decades, to test and learn, and then build out the connected infrastructure and analytics they’ve envisioned – with time in-between for inevitable course corrections. Flexibility is key.

“Something that will catch companies off guard is how quickly their competitors will be using IoT technologies and analysis,” says Covert. “A lot of people are thinking of this as something that’s off in the future, 3, 4 or 5 years. They need to be thinking more along the lines of 3, 4 or 5 months.”

Among the weighty decisions leaders face about Internet of Things, much attention seems focused on choosing the right hardware and software, and where to install them.

“The equipment choices are important, absolutely,” says Charlie Covert, Vice President of Customer Solutions at UPS. “But IoT’s true value will be found in the data those devices carry. But having a clear vision can not only reduce IoT infrastructure expense, it can lead to data insights that can help to differentiate you in the market.”

Covert shares an example of what UPS Chief Information Officer, Juan Perez, calls the company’s “Innovation Cycle.” In that cycle, the technology built to support operational improvements also helps to create new UPS® services and capabilities.

“Analysis of the rich data captured by connected devices throughout our organization led to UPS Package Flow Technology, which evaluates and optimizes every step in a delivery cycle – that’s from the way it’s routed and tracked to the way it is loaded, and then delivered,” Covert says. “The visibility provided by Package Flow Technology then led to breakthrough customer solutions, like Package Intercepts, the UPS Access Point Network and UPS MyChoice service."

UPS’s Jack Levis adds the groundbreaking ORION route optimization system to the list of innovations enabled by Package Flow Technology. Levis, Senior Director of UPS Process Management, has helped to pioneer the system that uses advanced mathematics to crunch millions of data points per second, and guides UPS drivers on the most efficient daily route.

Says Levis, “I think the first step for everybody is, get your data in order. Understand the data you use to run your business, and gather all the data you can and start analyzing it. Whether you're a large company or a small company, leveraging the data you already have into meaningful insights puts you that much further ahead of the game.”

Prepare for the Data Avalanche  

For some perspective on the volume of data IoT will generate, an MIT Sloan report notes that sensors embedded in GE machines alone can collect “50 million data variables from 10 million sensors.” And those numbers are surely set to skyrocket. Global projections for connected devices in use by 2020 range from between 25 billion to almost 200 billion. 

Says Covert, “When you look at the amount of information that will be collected from various IoT technologies, it can definitely be overwhelming. But now is the time to take that first step,” he urges. “Start talking about the data you need, and set out a phased approach for how you’re going to get there.”

Levis suggests that companies address fundamental needs as soon as possible. “It would be unusual for midsize companies to already have on staff all the expertise they will need to manage and make sense of their data. Outside expertise can be invaluable for basic steps, such as auditing the data sources you have, and how to access and store the data,” he says. “The right questions to ask at this point are, ‘Do we need rackspace?’ ‘What analytics tools will we need?’ and equally important, ‘What staffing and expertise will we need to maintain an ongoing plan for data gathering, analysis and feedback?’”

Start with the Problem, and Work Backwards

The late Steve Jobs, master of turning the seemingly impossible into multi-billion dollar products, once said, “You have to work hard to get your thinking clean to make it simple.” 
 
Not to imply that it will be simple to find data that will propel your company into the next century, but the sentiment is appropriate to IoT, because of the shift in mindset that will be required. For example, leaders facing big projects are conditioned to gather reams of facts in order to determine a course of action. Instead, what if you took a different path? What if you detailed the things you want to accomplish and solve, and then worked backwards to identify where the data you need resides – or where it will reside once you get access.

Not only can working backwards from help you narrow down the parts of your operation to connect, you’ll be ready to use the data that eventually comes streaming back at you – those bytes and bytes and bytes of data.
 
“Starting with the problem or opportunity keeps you from solving a lot of little problems that really have no impact,” explains Jack Levis, Senior Director of UPS Process Management, and a pioneer of the ORION optimization system.

Evolve Your Analytics Capabilities over Time  

The ultimate success of Internet of Things initiatives, like any enterprise effort, relies on certain mainstays. For one, the effort will need long-term corporate commitment and funding to maintain momentum. Second, it must be backed up with a documented architecture and implementation plan, to make sure it can be carried out seamlessly regardless of personnel changes. 

Levis suggests adopting a progressively sophisticated approach to their analytics plan, which has become an industry best practice. “The first step, usually thought of as descriptive analytics, reports what’s already happened, where am I today?”

The next phase applies more complex analysis and algorithms that can make the insights more predictive in nature, or Levis’ words, “Based on what’s happened in the past, what will happen tomorrow?.”   

In the next phase of this long journey, Levis explains that the analysis becomes prescriptive, which may suggest preventive actions, determining the optimal locations for new facilities, or for determining inventory levels that maximize order fulfillment rates, while reducing waste.

Sometimes, the suggested actions may seem counterintuitive. In the case of UPS’s ORION, the program uses accumulated data to determine the most efficient driver route, and gets steadily smarter as it learns. At first, many experienced drivers balked at ORION’s suggestions, for example, to pass three stops and return to them later. In the end, drivers learned to trust ORION, and the proof is in the 100 million fewer miles driven last year with the same number of stops.

“People tend to think about what’s best to do right now, and then the next step, and so on. But analytics tools like ORION think of an entire network. Any individual decision may seem wrong, but it’s best for the whole day.”

Talk. Think. Transform. (Repeat)

The first step in any major initiative is often the most difficult. But now is the time to start. 

Talk – Get the conversation underway. If you’ve already started, give it a jump-start by bringing an extended group of stakeholders to the table. Digitally connecting a supply chain, from upstream to downstream, affects literally everyone in your ecosystem. Leverage their knowledge, and let them leverage your knowledge for mutual benefit.
 
Think – Once you’ve brainstormed with your stakeholders, assign task forces and deliverables. Build a business case that secures executive and Board-level advocacy. Draw on outside partners for specialized expertise. Create the architecture; identify milestones and checkpoints.

Transform – As might be expected, this phase is not a “one and done.”  Most companies will take decades, to test and learn, and then build out the connected infrastructure and analytics they’ve envisioned – with time in-between for inevitable course corrections. Flexibility is key.

“Something that will catch companies off guard is how quickly their competitors will be using IoT technologies and analysis,” says Covert. “A lot of people are thinking of this as something that’s off in the future, 3, 4 or 5 years. They need to be thinking more along the lines of 3, 4 or 5 months.”

Learn more about the value UPS can bring to your IoT journey.


Download the high-res PDF of the article here