Military commanders have always lived and died by information—both quantity and quality. No surprise, then, that the US military has been an early adopter of the Internet of Things and is looking to expand its applications. But this new technology brings with it organizational and security challenges that present both opportunities and obstacles.
The turbulent seas of the North Atlantic in 1941 are a surprising place to find an explanation for the military’s need of the Internet of Things. But in that pitched battle between Allied merchant ships and German U-boats, information was the key to victory. Codebreaking and aircraft from escort carriers were the sensors, feeding information into centralized command centers, where decision makers on both sides routed submarine “wolf packs” or re-routed convoys of merchantmen. Victory went to those groups that could generate and analyze more information in a timely manner and, then, adjust not just their tactical posture but their logistical supply lines, intelligence groups, and support facilities.
Then and now, it is hard to overstate the importance of information to military commanders everywhere. While the rest of the world was waking up to the Internet in 1996, the US military, for example, was already outlining plans for “information superiority.”1 The military concept of decision cycles places information flow at the heart of all activities from logistics to intelligence; in some cases, information’s importance and impact is so great that it is classified in the same category as artillery—as a deadly long-range weapon.2 With information so central to all activities, the military is naturally hungry for technology or tools that improve communication, routing, or processing of information.
The Internet of Things (IoT) is one such technology. Whether called the Internet of Everything, machine-to-machine, or ubiquitous or embedded or ambient computing, the IoT is fundamentally about connecting disparate objects into larger networks.
While the military has been a driver in connected and machine-to-machine communications such as radio frequency identification, more commonly known as RFID, it has been slow to adopt true IoT applications that knit these communications into interoperable, automated cycles. Communications remain within their given channels, not easily shared or aggregated.
The challenge is that defense leaders wishing to take advantage of the IoT face a complex technological and regulatory landscape that threatens to mire their efforts in endless choices and challenges. This article aims to help leaders navigate complex IoT decisions by pointing out which applications may be better suited for their goals related to cost efficiency and/or warfighter effectiveness. In each case, we will use Deloitte’s Information Value Loop as an analytical framework to identify the key investments necessary to realize the IoT’s potential benefits.
The suite of technologies that enables the Internet of Things promises to turn most any object into a source of information about that object. This creates both a new way to differentiate products and services and a new source of value that can be managed in its own right.
For information to complete the loop and create value, it passes through the loop’s stages, each enabled by specific technologies. An act is monitored by a sensor that creates information, that information passes through a network so that it can be communicated, and standards—be they technical, legal, regulatory, or social—allow that information to be aggregated across time and space. Augmented intelligence is a generic term meant to capture all manner of analytical support, collectively used to analyze information. The loop is completed via augmented behavior technologies that either enable automated autonomous action or shape human decisions in a manner leading to improved action.
Military leaders, as with those atop any organization, use structure in order to understand the choices before them. The value loop (see sidebar) is exactly one such structure, providing the context to sort through all of the noise and determine where to apply the next investment dollar to bring value to the organization.
Military-oriented readers can recognize the foundation of the value loop as a decision cycle, with stages that echo “observe-orient-decide-act.” What is unique to the loop is its ability to simplify IoT decision making. The value loop is able to bring order to the menagerie of IoT technologies by showing how each fits into larger processes and supports decision making. But with the proliferation of IoT applications—each one promising to change the world—simply having a taxonomy of technology is not enough to help decision makers. Rather, the value drivers help to illustrate how the flow of information creates value. By understanding differences across these drivers between competing IoT applications, a defense leader is on the road to being able to determine what is right for his organization, what is transformative, and what is merely hype.
The first step is to determine exactly what the organization needs to accomplish. In the age of sequestration and the global war on terror, the overarching theme commonly seen in defense budgets is “seeking a balanced force.”3 The United States needs a flexible force with more combat power than ever to counter diverse threats from Syria to the South China Sea, but also one that is cost-effective and efficient in its use of resources. To meet the challenge, commanders naturally seek options that will either reduce costs, freeing up time and assets for core mission activities, or directly improve those activities themselves. We can categorize IoT applications according to the same logic: those that aim to improve cost efficiency, those that aim to improve warfighter effectiveness, and rare cases that aim for both.
While some aspects of the military are decidedly unique, other functions closely mirror their civilian counterparts. In these areas, such as asset tracking and facilities management, leaders can simply import existing civilian technologies to gain the advantages of new IoT applications.
Cost reductions through asset tracking: Many individuals with even loose military associations have experienced the tedium of filling out handwritten receipts. And all too many leaders know the more intense pain of long hours spent inventorying equipment by hand, reading off the serial number of each piece in turn. More than a mere inconvenience, these dated asset-management practices compromise the US Department of Defense’s (DoD’s) overall supply-chain effectiveness. A recent survey of DoD supply logistics managers identified ineffective data management as the primary risk to their supply chains.4 This lack of information directly results in equipment shortages on the one hand and waste of excess equipment on the other.
The DoD recognizes these inefficiencies and has long worked to improve them. As far back as 2005, the Defense Logistics Agency argued for the military’s adoption of RFID as a standard for supply-chain tracking.5 The potential benefits seemed obvious: better awareness of equipment location and status; faster, more accurate deliveries of needed supplies; and, of course, less manpower wasted on dreaded inventories.
However, even as an early adopter, the DoD has struggled to win widespread acceptance for RFID. Largely this is because the department has had difficulty demonstrating the return on investment for the subordinate commanders who bear the time and cost burden of implementing RFID.6 Without a clear picture of how they will benefit, leaders are reluctant to invest scarce resources in these technologies—an important point for IoT applications in general.
In terms of the value loop, this lack of investment stifles the loop where it begins, at the create phase. Without putting sensors on objects and connecting them, data are not created, and no information flows around the loop. Although the initial costs may seem daunting, mid-size defense organizations such as the Robotic Systems Joint Project Office have been able to implement asset-tracking systems for as little as $400,000.7 Similar applications in the civilian sector have been able to recoup their investment within the first year.8 By applying location sensors to moveable patient monitors, seven hospitals across the nation were able to more efficiently use the monitors. One hospital saved more than $500,000 a year in the cost of buying or renting new monitors. More importantly, when integrating sensors into a larger tracking system, hospitals were better able to map the flow of patients, resulting in maximized bed use and annual revenue increases in the millions of dollars.9
Cost reductions through facilities management: Civilian IoT successes in utilities and facilities management also can provide a useful roadmap for the military, in an area with immense potential for value capture. The DoD holds the largest US portfolio of facilities and is the nation’s largest single energy user.10 In FY 2016, the department will spend more than $10 billion to maintain and repair those facilities.11 In an era of tight budgets, facilities maintenance and utilities are often easy targets: The Marine Corps alone was forced to cut $7 million from its utilities budget for 2016.12 Since turning off the lights on the troops is not a viable solution, improving energy efficiency can be key.
An IoT application already proven in the civilian world may offer promise for the military’s energy-efficiency goals. In the civilian world, the IoT has enabled centralized building-management systems to target costs by monitoring and coordinating utilities and building functions. For example, the Central Bucks School District in Pennsylvania has saved $15 million in energy costs in five years by implementing an IoT-enabled facilities management system.13 These systems go beyond simply dimming lights or heating only occupied rooms—they include continuous monitoring, which identifies hard-to-detect inefficiencies. One application at Western Kentucky University found unexpectedly high energy usage in off-peak hours via leaks in the air-handling system, causing major waste and expense.14
As with asset tracking, the bottleneck in the Information Value Loop of facilities management starts with create: Useful data are simply not being created. With existing, proven technology, defense leaders can more easily act to see desired savings. A pilot study implementing IoT solutions at Great Lakes Naval Station combined real-world weather data, energy consumption, comfort thresholds, and data collected from buildings into a machine-learning algorithm designed to reduce energy consumption. The study found reductions in energy usage of 20 to 30 percent, suggesting that, if implemented across the DoD, annual savings of $500 million could be possible.15
Challenges to adoption: Why has the military struggled to implement cost-reducing IoT applications that for-profit companies have used to great effect? The challenge in implementation is not a matter of technological lag—the devices and capabilities are already in use in the private sector and have demonstrated effectiveness in specific military installations. Rather, slow IoT adoption can be explained by analyzing the fundamental structural and cultural differences between the private sector and the military.
Military leaders, in contrast to their civilian counterparts, are rarely able to keep any savings they realize from implementing facilities or utilities efficiencies—they may even be penalized. The difficulties in moving funds between different appropriations means that a commander who saves $1 million in energy costs is likely unable to use that same $1 million to buy more equipment or hold more training exercises.16 Instead, the money may be “lost,” with commanders often suffering a reduction in the next year’s budget as a reward for their thrift.
The possibility of misaligned incentives suggests a need for a centralized management approach: one organization that is able to design, test, and implement energy-saving measures, then mandates the successful efforts across one of the five US military services or even across the Joint Force. If that organization also managed the facilities or utilities budget as a whole, it would be able to realize and allocate the savings created by energy efficiency, aligning incentives. With the creation of organizations such as the Air Force’s Installation and Mission Support Center, all the services seem to be on the path toward centralized management. Such organizations are well positioned to capitalize on demonstrated civilian success in IoT facilities management and begin to reach the cost-reduction goals of the modern military.
Where civilian IoT applications could deliver efficiency and cost reductions in supply-chain tracking and facilities management, the unique demands of warfighter support require military-specific IoT applications.
Consider Julius Caesar commanding the famed Tenth Legion at the battle of Sabis in 57 BC. Briefly surprised by the Gauls, Caesar rode to the front lines, observed his forces, and shouted orders to “each Centurion by name.”17 For centuries, that was the model of command and communications: Battlefield commanders “pulled” information about friendly and enemy situations from the dust of the melee and then “pushed” commands and orders down to the tactical units. For Caesar, with all of his forces easily seen and within shouting distance, this push-pull model was fairly straightforward, but the expansion of modern battlefields has introduced new challenges.
Today the pull is often accomplished by data fusion, trying to give a commander the widest, most diverse picture of the battlefield. The push is the challenge of operational communications: how to disseminate orders to and among tactical units. The military has already implemented many of the foundational components of IoT in both pull and push; however, data often remain disconnected—separate value loops with separate flows of information. Continued development in network technology and data standards promises to create a tactical IoT that can unify the push and the pull, remaking battlespace awareness into a truly modern process—one that Caesar would not recognize.
“Pull” data fusion: Informed decision making is predicated on having comprehensive knowledge of the battlefield: reports from a range of locations, taken over a span of time that paint an accurate picture of the situation. For Caesar, this was as simple as looking across the field of clashing warriors. But battlefields, and the sophistication of the battles fought on them, have grown. The information a commander needs to make effective decisions has expanded exponentially, meaning that commanders often bring together volumes of diverse data to understand their battlespace. In terms of Deloitte’s IoT framework, they focus on aggregate.
The importance of data in modern warfare poses two distinct challenges for a commander: handling the sheer volume of data produced, and integrating numerous types of data into one coherent battlespace picture.
Storing and (quickly) retrieving large volumes of data is not a uniquely military task—other areas of government have made substantial progress on this challenge. With an estimated 220 exabytes of data worldwide to be stored in 2015, a civilian-world solution has been to move data to the cloud.18 Defense and intelligence leaders have followed suit: The CIA and Defense Information Security Agency (DISA) have leaned on civilian expertise, working with commercial companies to bring the cloud and software to secure government networks.19 Thus, the infrastructure for dealing with the data volume of tactical IoT applications is, potentially, already in place.
Diversity of data, on the other hand, poses unique challenges to cloud implementation in the military and complicates private-sector comparisons. Drones are a particularly illustrative example: The Department of Defense has fielded at least 13 different types of unmanned aerial systems, each with an array of sensors producing multiple types of data.20 This doesn’t begin to address the issue of compatibility—just one of those data types, video, can come in more than 20 different file formats, with even more choices of encoding and frame rates.21 (YouTube—a civilian site that hosts only videos—can support just nine file formats.22) Indeed, military data-fusion applications incorporate not only videos but still imagery, signals intelligence, human intelligence, ground sensors, battlefield reports, map data, and a host of other data sources.
As the value loop shows, aggregating these disparate types of data necessitates a common set of data standards. With multiple agencies, commands, and military services involved in the production, transmission, and consumption of all of these data types, creating a common set of standards will likely require a senior-leadership mandate designating an executive agent such as DISA to adjudicate what the standards should be.
“Push” operational communications: As with data fusion, the military has a long history of working with connected communication in operational communication. Where the distributed communication of smartphones has driven civilian IoT development, every soldier since the 1940s has carried a radio.23 Where real-time traffic-monitoring apps began revolutionizing our daily commutes in 2009, Blue Force Tracker has been an everyday part of ground combat for the better part of two decades.24 Despite communication technology’s importance to the fabric of military culture and communication, however, it has not evolved into true IoT functionality, with data that are open and discoverable between systems. Information bottlenecks are frequent within operational and intelligence networks. The typical Brigade Combat Team communications network in 2013 featured 5,000 discrete nodes communicating over more than 18 distinct network types.25 Where data do cross between these networks, they often do so via an arduous manual process.
The dream of true IoT capability in operational communications is not new. As far back as the 1993 launch of the Army’s now-defunct Land Warrior program, the military has tried to integrate frontline forces’ sensors, weapons systems, and communications. Until recently, the technology simply did not exist to make these systems work as desired: They could not connect to other systems such as Blue Force Tracker, nor did the communications systems have enough power to transmit the volume of information created.26 To put this in terms of the Information Value Loop, there was a bottleneck at the communicate stage.
Where “pull” was limited by the process challenge of breaking down multiple, siloed data standards (i.e., failures in aggregation and the standards that support them), IoT usage in operational communications is constrained by the technical limitations in mobile communications networks’ bandwidth and robustness. While a consumer’s 4G LTE smartphone will routinely post download speeds in the range of 8 to 9 Mbps, the military’s commercial satellite network used for mobile network access posts a top speed of less than 0.5 Mbps.27 These speeds are more than enough if soldiers need only voice communication or to send short text-only messages, but current military communications systems cannot provide a soldier in the field with the bandwidth that a true IoT application would require—and certainly not wirelessly.
While the military can, to a certain extent, ride the wave of civilian mobile telecommunication such as 4G LTE, those advances will likely need to be paired with advanced, military-specific communications architectures. After all, the average consumer does not need a network of rugged, encrypted, frequency-hopping, multiband radios. For the military, though, research in this area is beginning to bear fruit. For a number of years, DARPA has been experimenting with “mobile ad hoc networks,” designed to form a self-creating and self-healing mesh of communication nodes, with setup time measured in minutes instead of days. DARPA envisions networks of more than 1,000 nodes providing individual soldiers with streaming video from drones and other sensors, radio communications to higher headquarters, and advanced situational awareness of other soldiers’ location and status.28 Some of the products of such research are already reaching the battlefield: The Army has begun testing prototypes of its Integrated Sensor Architecture, which allows for dynamic discovery of sensors.29 Using this architecture, for example, a soldier walking through an area could quickly locate a sensor hidden in the ground and read off data about whether any enemy vehicles had passed through the area over the past 24 hours.
Without the ubiquitous cellular signal upon which we rely in our daily lives, these military IoT networks operate over tactical radios. The next generation of high-bandwidth radios that could make these integrated networks a reality are already under development. Air Force Special Operations Command recently released a contract solicitation describing requirements for a tactical radio that sounds more like a futuristic smartphone than a traditional single-channel VHF radio. The Special Operations Forces Multi-Function Radio must be able to form a 100-node self-healing mesh network and automatically connect within five seconds. With a minimum 5 Mbps data rate, it is about as fast as the 4G LTE smartphone in your pocket, while meeting military durability and encryption requirements.30
Unified battlespace awareness: With these advances in communications architecture and devices, a true tactical IoT will be nearly here. The high data rates and flexible communications architectures now reaching operating forces do not simply improve upon existing communications—they have the potential to change how soldiers in the field shoot, move, and communicate. Individual applications already allow a soldier to view streaming video from unmanned aerial vehicles overhead, see nearby soldiers’ ammunition status, text message a “call for fire” to supporting artillery, and even allow headquarters to look through a rifle scope. The current challenge is integrating all of those applications.
Indeed, all of this is about more than simply giving every soldier a smartphone—it is about, for the first time, unifying the push and the pull of command-and-control. Where now commanders pull data from the front lines and push orders down, a tactical IoT would allow for organic flow of information up and down the chain and around the value loop, with both the commander in the tactical operations center and the soldier on point enjoying increased situational awareness. The tactical military IoT is the next step in command-and-control—and the first step toward a new approach radically different than Caesar shouting to his centurions.
We have already seen how repurposed civilian IoT applications can help the military cut costs, and where military-specific IoT applications in battlespace awareness can increase warfighter effectiveness, but there is one area where a revolution is pending with the potential to do both: autonomic logistics.
Like the IoT, autonomic logistics goes by many names: predictive, proactive, prognostic, or condition-based maintenance. But in each case, the concept revolves around the ability to use real-time data about usage and system performance to predict failures. Also like the IoT, this concept is not new, dating back to DARPA-sponsored research in the 1970s, but has come into its own only with the inexpensive computing, sensor, and communication technology of the past decade.31
Meet ALIS: The most advanced example is the Autonomic Logistics Information System (ALIS, pronounced “Alice”) of the F-35 joint strike fighter. ALIS uses sensors embedded throughout the aircraft to detect performance, compare to parameters, use sophisticated analytics to predict maintenance needs, and then communicate with maintenance staff so that the right parts are ready when needed. This represents a quantum leap over standard maintenance practice, in which a maintainer’s only guide is checklists based on how many hours an airframe has flown.
Statistical experiments have demonstrated improvements in both cost and aircraft availability over these standard maintenance methods.32 These improvements occur at every link of the maintenance supply chain: Ordering the right parts at the right time means a simpler logistical train; less unnecessary maintenance means fewer maintenance personnel and lower costs; and with fewer aircraft grounded for maintenance, more are available for flight, increasing potential combat sorties. According to William Scheuren, the former head of the DARPA program that eventually brought forth the F-35, the goal of autonomic logistics is to reduce the complexity of the logistical train by 50 percent, reduce the number of maintenance personnel by 20 to 40 percent, and increase the number of combat sorties by 25 percent.33
While logistics is everywhere, like battlespace awareness, ALIS is a uniquely military application. Civilian manufacturers, for example, can use networked sensors to monitor the performance of engines mounted on commercial airliners and to suggest maintenance to ground crews, but no system monitors all aspects of a vehicle, from engine to avionics, or integrates all aspects of the supply chain, from the aircraft itself to maintainers to parts depots.34 This project’s scope and scale are such that the military is breaking new ground and, naturally, uncovering new challenges along the way.
One of those new challenges implies that incorporating IoT technology can be something of a cautionary tale. The development of the F-35 has seen numerous cost and schedule overruns: Already the most expensive weapon system in history, the F-35’s sustainment costs are set to be nearly double those of the F-15C/D, F-16C/D, AV-8B, and F-18A-D combined.35 And a large part of that sustainment bill is due to performance issues and cost overruns within the ALIS program.
The complex integration of hardware and software found in ALIS has created a number of technical conflicts and glitches; in some cases, these faults have actually weakened the very metrics ALIS was designed to improve. In one documented case, ALIS incorrectly signaled that the landing gear of an aircraft had failed when in fact the fault was with a small component near the gear.36 This error forced maintainers to repeatedly inspect the landing gear, forcing the grounding of an otherwise healthy aircraft. In this case, the sensor worked properly in detecting a fault, but the analytical software failed in understanding its importance. Marine Corps Lt. Gen. Robert Schmidle hinted at the root of the problem in a recent interview: “[W]e need to have the ability to override the algorithms that are built into that system to determine whether an aircraft is safe to fly or not.”37
To put this in terms of our value loop, what Gen. Schmidle and the F-35 team are discovering is that as the analytics or augmented intelligence within an IoT application become more complex and independent, new, previously unidentified issues emerge—specifically, the boundary between humans and augmented intelligence (AI) requires careful design. In the case of relatively simple augmented intelligence—fitness bands, for example—the human-AI boundary is relatively simple as well: It issues an alert that the user is free to follow or ignore. But as AI grows in complexity, it cannot easily communicate all the parameters to a human user, forcing the system to take certain pre-programmed actions. However, if those pre-programmed actions are ill chosen or inflexible, they can lead to unintended negative consequences—such as the grounding of perfectly good aircrafts.
In this respect, the F-35 is on the IoT’s cutting edge. Previous Deloitte research concludes that even the most advanced AI applications available today still require a “human in the loop” in order to achieve promised levels of accuracy. How exactly to combine human judgment and AI computational power in the right way remains a question for each new AI application.38
Truly revolutionary military IoT applications of the size and scope of ALIS will continue to reveal challenges at the boundaries of our knowledge of cognitive and computer science. While the ultimate solutions to these issues likely lie in laboratory research, the military credo of “improvise, adapt, and overcome” will likely be the near-term fix. As the Marine Corps prepares to declare initial operational capability of its F-35 fleet, simple fixes such as a manual override will likely serve as an interim remedy until more robustly engineered software fixes are available. Complex problems with simple interim solutions will likely form the pattern for advanced IoT applications moving forward.
With strategy concepts such as “net centric,” “information dominance,” and the emergence of cyber as an entirely new domain of operations, information always has and will remain central to the military’s efficiency and effectiveness. Naturally, IoT technologies and architectures that are designed to move and process information more quickly and in distributed environments seem like natural fits for military applications. But the IoT can be a highly technical and complex landscape to navigate, and adapting applications for military purposes adds layers of requirements and specific uses. All of this can make it difficult for leaders to determine which IoT applications are right for their organizations and how best to implement them. Our analysis has shown a number of commonalities across different IoT applications that can help organizations move quickly toward technological transformation.
Like all planning, the first step is to decide what you need to achieve. In a world where every defense organization faces twin threats of budget cuts and dangerous enemies, the first choice is whether IoT applications should aim to achieve cost reductions, increase warfighter effectiveness, or both.
For the military as in any field or industry, there is no one-size-fits-all solution to the IoT. The key is to take a reasoned approach to investigating IoT applications. Start from your mission need, whether cost reduction, warfighter effectiveness, or both, and look forward from there. Battlefield objectives have shifted in the centuries between Caesar and the Islamic State—there’s no reason not to use all available tools and technology to achieve those objectives.
Deloitte’s Internet of Things practice enables organizations to identify where the IoT can potentially create value in their industry and develop strategies to capture that value, utilizing IoT for operational benefit.
To learn more about Deloitte’s IoT practice, visit http://www2.deloitte.com/us/en/pages/technology-media-and-telecommunications/topics/the-internet-of-things.html.
Read more of our research and thought leadership on the IoT at http://dupress.com/collection/internet-of-things/.