John Rossman’s new book, The Amazon Way on IoT.
Amazon has long been at the forefront of emerging technologies, and the Internet of Things (IoT) is no exception. IoT technology, which allows objects to send and receive data through internet connectivity, is changing the way we interact with everyday objects. For example, you can control your air conditioner through a mobile app or track your health stats through a wristband.
On the consumer side, Amazon introduced the Dash Button, which allows customers to reorder an item with the touch of a standalone button. For businesses, the eCommerce giant launched AWS IoT, a platform for developers to build, manage, and analyze their own IoT devices.
John Rossman, author of The Amazon Way on IoT.
In his new book The Amazon Way on IoT, former Amazon executive and renowned thought leader John Rossman reveals strategies for implementing IoT technology through the lens of Amazon’s pioneering approach. Below, read an exclusive excerpt in which Rossman reveals how utilizing connected devices can drastically improve business by decreasing waste and costs.
Looking to dive into the world of IoT? Read our interview with Rossman to learn what IoT technology can do, why you should have your eye on it, and how to launch a career in this cross-functional field.
Join us on campus to meet Rossman and hear him talk about the strategies every leader needs when it comes to IoT. Rossman will be speaking on January 9 at our Orange County campus and January 10 in Downtown Los Angeles.
***
Chapter 3, Relentless.com—Continuous Improvement via Connected Devices
One of the more unique aspects of working at Amazon is that everything—every process, every customer experience, and every function—has an improvement plan and roadmap. Compare that with the typical company, where, other than the occasional reorganization, processes that aren’t broken stay largely the same from year to year.
You can find nods to this focus on continuous improvement all throughout Amazon’s leadership principles and history. Bezos originally named his company Relentless.com. In fact, if you type “www.relentless.com” into a browser, it will still take you to Amazon.com. While he ultimately decided against such a literal moniker, “relentless” still perfectly encompasses Amazon’s nature. The company is dead set on constantly exploring and reinventing itself through key leadership principles and an undying belief in the power of technology.
For Amazon, IoT is the latest big opportunity to stay relentless.
Principle 3: Connected devices are a powerful enabler for monitoring and improving your operations to make your company more efficient, competitive, and profitable.
In this chapter, we’ll explore the tools Amazon and others use to think about ways that connected devices can reengineer a company’s current processes and capabilities to improve quality, decrease waste, reduce cycle times, and, as a result, decrease costs.
Amazon’s dedication to continuous improvement is a key part of company culture.
Even the teams and team leaders that aren’t sure whether their projects will get funded have a deep understanding and written articulation of how their idea will scale if it does. That focus on how to do more with less means that projects that are funded can hit the ground running. Leaders already know how a project will improve the customer experience and how it will reinforce Amazon’s flywheel. The Amazon flywheel is the strategy Bezos created to explain how customer experience would reinforce its larger business goals and vice versa. (We’ll talk more about that in future chapters.)
This company-wide expectation is reinforced by things like Amazon’s evaluation process, which assesses employees for, among other things, their commitment to continuous improvement. “Always looks for ways to make Amazon.com better,” the standard reads. “Makes decisions for long-term success. Investigates and takes action to meet customers’ current and future needs. Not afraid to suggest bold ideas and goals. Demonstrates boldness and courage to try new approaches.”
Amazon’s approach to process improvement illustrates the fact that there are really two kinds of innovation. “Big I” innovation is what comes to mind for most people when you talk about innovation. It generally leads to a new product or feature or a completely new experience. In the best cases, those experiences feel like magic. (Think about drone delivery or PrimeNow’s two-hour delivery program.)
But there’s another type of innovation that’s just as—if not more—important to a company or product’s success. And, as Amazon has demonstrated, it’s one of the most consistent ways to take your products, services, and processes from good to great. This kind of innovation is focused on the continuous refinement and improvement needed to create frustration-free customer experiences. I call it “little i” innovation, commonly known as continuous improvement.
Of course, Amazon is just one of many companies that have found value through a focus on continuous improvement. It’s likely you’re at least familiar with one or more of the business methodologies it has inspired.
- Lean—the philosophy of creating more customer value with fewer resources.
- Toyota Production System—a management approach intent on eliminating all waste, which includes key strategies such as “Just in Time” inventory and demand signals.
- Statistical Process Control, or SPC—a system of attaining and maintaining quality through statistical tools that emphasizes root-cause elimination of variation.
- ISO 9000 Quality Management—a set of quality certification standards based on eight management principles, including continuous improvement and fact-based decision making.
- Six Sigma—a data-driven methodology for eliminating defects, reducing costs, and eliminating waste.
All of these strategies empower employees at the companies that use them to gather data and to act on the insights that data provides. They are encouraged to drive change and improvement from within. But each of these strategies was also created before IoT.
The introduction of ubiquitous connected devices has changed the rules of the data game, creating the possibility for real-time feedback loops that power continuous improvement programs.
Instead of living in a world of manual data collection, which creates limited, slow, and stale data sets, IoT creates an exponential stream of affordable real-time data. That flood of data empowers companies to focus on the continuous improvements to their internal systems, saving them time and money while increasing productivity and consistency.
How Amazon Took Operations from Good to Great
Today Amazon’s operations—the way they fulfill, ship, track, and deliver your orders—are world class. But they didn’t start out that way. Amazon measured, refined, and executed its way to greatness. It embraced continuous improvement as a way of life.
By building that dedication into its company culture, creating an operational-improvement heritage, Amazon has been able to build out consistently high-quality, low-cost facilities all around the world. They now boast three hundred fulfillment centers across fourteen countries.1
This kind of consistency gives Amazon the confidence and competence to guarantee incredible service: Amazon Fresh, Amazon’s home grocery delivery service, lets customers schedule delivery within a fifteen-minute window. That kind of customer service takes incredible forecasting and execution capability—ability built on the back of Amazon’s world-class supply-chain heritage.
That level of precision wouldn’t be possible if Amazon hadn’t made a concerted effort to take advantage of connected devices and the data they provide.
In the early 2000s, the leaders of Amazon’s fulfillment and operations capabilities decided to implement Six Sigma, a data-driven five-step approach for eliminating defects in a process. Define, measure, analyze, implement, and control—or as it is referred to in Six Sigma, DMAIC. This is the root improvement cycle in Six Sigma and sets up the methodical, measured steps and mindset to squeeze out defects, costs, and cycle times.
Six Sigma was introduced by Bill Smith, an engineer at Motorola, in 1986. In 1995, Jack Welch used it at General Electric to much success. The term itself is used to describe a manufacturing process that is defect-free to six standard deviations. In other words, the process is 99.9996 percent accurate.
One of the challenges of completing a Six Sigma initiative is that so much of the effort—generally up to 25 percent—lies in collecting data. Depending on the project, manual data collection can be not only difficult but inaccurate. The data itself is often of questionable quality, skewed by bias or cut short due to time and effort.
Because of these challenges, Six Sigma certifies professionals in a set of empirical and statistical quality-management methods to help them execute on the process successfully. These professionals are installed in an organization during a Six Sigma process to make sure everything is completed successfully.
There are several levels of Six Sigma certification, but the most involved is called a Black Belt. Black Belt practitioners have received significant training and are deeply vested in applying Six Sigma, dedicating 100 percent of their time to its application. It takes a certain kind of person to be a great Black Belt. Black Belts are generally nimble problem solvers, good project managers, and facilitators. They are crafty at collecting data and have a strong background in statistics and math.
As you can imagine, these kinds of people are also highly sought after and well compensated. Creating a team of Black Belts within your organization is one of the biggest cost drivers of Six Sigma initiatives.
That’s where IoT comes in.
Using connected devices to collect data frees up the Black Belts in an organization to tackle more projects. It also leads to faster Six Sigma initiatives and a much richer, more reliable data set.
Connected devices can bring visibility to your company’s operating conditions, giving you real-time insight into the flow, status, and state of key items in your process. Not only does this enhance your understanding of needed improvements, but it builds a way to scale operations with active quality and measure built into the process.
At the time Amazon integrated Six Sigma into its operations, the company was experiencing a disconnect in a process it calls SLAM. SLAM stands for the ship, label, and manifest process. Every time something, like a printer, is ordered on Amazon, that printer is placed in a box in an Amazon fulfillment facility, labeled, sorted, and shunted through the fulfillment center, until eventually it’s placed in an outbound truck. That’s the SLAM process. At peak, Amazon ships over one million packages a day.
When Six Sigma was introduced, packages were labeled and moved down conveyor belts before being manually sorted and delivered to the correct docking station. This worked well most of the time, but there was no final confirmation that the package had actually made it onto the right truck, and there was no visibility—for the company or the customer—about where exactly a package was in the outbound process. As a result, packages were occasionally missorted.
An occasional missort doesn’t sound like a big deal, but over the course of a year, missorts can cost a company like Amazon millions of dollars. More importantly, even one missort breaks Amazon’s underlying promise to its customers: the promise that all of their orders will arrive in their hands on time.
For Amazon, the solution was to create a positive automated confirmation, or “visibility,” that a package had moved correctly through all logistics checkpoints after its shipping label had been applied. The change was simple in concept but incredibly complicated in implementation.
To execute, Amazon installed sensors and readers across its conveyor system. The sensors would automatically scan a package’s barcode as it moved through the SLAM process. Since packages were scanned to destination-specific staging areas, the sensors allowed Amazon to track the whereabouts of specific packages at any given time in the SLAM process. Furthermore, as Amazon employees loaded those packages into the outbound trucks, scanners on the bay doors would alert them if a package was about to be loaded into the wrong truck.
By creating a positive-confirmation system for its packages, Amazon lowered its missorts to within Six Sigma’s 0.0004 percent accuracy range. That’s fewer than four packages missorted in every million.
In Amazon’s SLAM process, the IoT answer was relatively complicated to implement. In many continuous improvement cases, though, IoT-based solutions can be exceedingly simple.
For example, if you run a corporate facility, in many states you are required by law to schedule monthly inspections of every fire extinguisher in that facility and file a report. It’s an incredibly important rule from a safety and liabilities perspective, but it’s also a total pain in the butt to schedule and complete. As a result, companies often miss a month here or there. Enter en-Gauge, a company that sells connected fire extinguishers. en-Gauge’s solution alerts a building’s security system if an extinguisher becomes noncompliant. It will also notify a facility manager if an extinguisher is pulled from its mount. The savings are not only in manual labor but in the accuracy of the inspection.
Integrating IoT-Driven Continuous Improvement Into Your Operations
You may be struggling to imagine how connected devices could empower and supercharge process improvements in your company. That’s OK. There are several questions that you can ask yourself to help identify situations that might benefit from an IoT-driven continuous improvement process.
1. What operating condition information would be valuable to your company? Let’s say you operate a truck fleet. One of the largest operating costs and safety issues in trucking is tire wear and maintenance. Tires are expensive and wear out relatively quickly because of the weight of the load they carry and the long distances truckers travel. A punctured tire can lead to lost goods, but it’s also the cause of many accidents. Sensors and real-time data about tires could help a driver or a trucking company monitor tire wear, tread, and air pressure to avoid acute tire compromise.
2. What manual data entry or logging is done in your business today? Across the healthcare industry, a huge amount of data is being collected about patients every day. Nurses measure and record patients’ temperatures, blood pressures, and medications, among many other things. Using machines that automatically record that data in a patient’s file would not only save money but would mean fewer errors made and more lives saved.
3. What is the incomplete and inaccurate data in my business? Imagine you run a sales team. You’d like to be sure that that team is out making a certain number of visits or stops every day, but tracking every sales associate—or even asking them to track themselves—is a tremendous burden. There’s also a huge variety in the accuracy and amount of data that different teams may choose to input into a company’s CRM system. Where did each salesperson go? How long did that person stay? How many outbound calls did an associate make? All of this could be captured with sensors and devices.
4. What inspections and audits are done today? en-Gauge, which we mentioned above, is one example, but any kind of regular inspection or internal audit can be a great candidate for automation. Picture a storage bin full of Kindles in an Amazon fulfillment center. Simply adding a weight sensor to the bin (otherwise known as a scale) would allow the company to consistently reconcile its physical inventory with what’s recorded in its inventory system.
5. What shrinkage, damage, or underutilization occurs in the business? Hospitals are a great example of a situation where equipment is costly, mobile, and critical. Knowing where critical equipment, like a crash cart, is and whether it’s ready for use is mission critical in a hospital. Real-time location systems, or RTLSs, is an entire category of IoT solutions providing information about the whereabouts and operating state of medical equipment.
6. What are the operating risks? Knowing—and tracking—key operating risks can save millions of dollars and help you avoid catastrophic risk. The oil and gas industry is aggressively adding sensors into their pipeline systems so that they can have real-time information about operating risks or weaknesses in pipeline systems. PG&E, which operates tens of thousands of miles of pipeline, even taps into data from fire, transportation, and civil emergency-response systems to identify risks. 2
7. What are the quality issues and drivers of customer contacts? Think about what information would provide insights into the root causes of problems with your product that your customers might experience. What sensor would identify or create an indication of the situation? As a customer, I don’t know if I have a furnace or AC issue until it happens or unless I schedule a routine maintenance check-in. A sensor installed in those machines could remind you that it’s been six months since you switched your air filter or that there’s a 75 percent probability that a part will fail within the next three months.
Special Snowflake? Think Again
I was recently challenged by a potential client who felt that his company was already world class in its operations. There really wasn’t much opportunity, he said, for ongoing operational improvement through the Internet of Things. I told him the story of General Electric.
GE is renowned, and often copied, for its operational excellence. The company is widely considered a leader in the adoption of Six Sigma across many aspects of business. In spite of all this, GE sees operational improvement through IoT as a monumental opportunity. In 2015, GE announced that it would be rebranding itself as the “Digital Industrial Company,” using IoT to drive improvements not only in its operations but to create new businesses. As CEO Jeff Imelt wrote in his 2015 annual letter, “We accelerated our transformation as a leader in the Industrial Internet, becoming a ‘Digital Industrial’ company. In the Industrial Internet we see the next great wave of productivity—both for our company and for the customers we serve. We are a company that invests in broad industrial transitions, and they don’t come much bigger than the full application of data and analytics to machines and systems.”3
If GE is betting the business on IoT to drive the next wave of operational improvement, I’m betting that your company can find ways to improve even world-class capabilities.
The trick? You just have to be relentless.
1 Marc Wulfraat, “Amazon Global Fulfillment Center Network,” MWPVL International, August 2016, http://www.mwpvl.com/html/amazon_com.html.
2 Alex Jablokow, “How the IoT Keeps Oil and Gas Pipelines Safe,” PTC, November 3, 2015, http://blogs.ptc.com/2015/11/03/how-the-iot-helps-keep-oil-and-gas-pipelines-safe/.
3 Jeffrey R Immelt, “GE 2015 Annual Report,” GE, February 26, 2016, http://www.ge.com/ar2015/letter/.