Search:  

 Blog     e-Newsletter       Resource Library      Directories      Webinars     Apps     EBooks
   AFCEA logo
 

Predictive Insight Offers Commanders Edge

January 2007
By Lt. Gen. Jack Woodward, USAF (Ret.), and Ryan M. LaSalle

 

The era of network-centric warfare combines precision operations, agile ground forces, unprecedented surveillance and real-time communications that connect diverse forces in remote locations. A new technology-based approach may consolidate available data in a way that allows commanders and logisticians to predict outcomes mathematically.
©Ed Kashi/Corbis

Military leaders can peer over the horizon to gain significant tactical advantage.

Military commanders looking for a battlefield advantage that can tip the balance dramatically in their favor may be able to benefit soon from a promising new technology application. Called predictive insight, it holds the potential not just of making the concept of complete battlespace awareness a reality but also of taking that concept a giant step further.

Predictive insight offers the possibility of even more than an up-to-the-moment, 360-degree look at all combat conditions. It could provide the information that military leaders need to predict with far greater accuracy what will happen next on the field and to gain an immeasurable advantage over the enemy—and possibly help lead to victory.

Military personnel operate in a theater of constant uncertainty. They must determine rapidly whether a car speeding toward a checkpoint is a suicide bomber or a civilian rushing his pregnant wife to the hospital. Or, they must try to predict whether the next attack is most likely to come on a military outpost. Often, these personnel are forced to make split-second decisions in the face of numerous unknown factors.

Predictive insight tries to change the odds to deal with that uncertainty better. It works by taking advantage of technological capabilities that already exist and exploiting them to a greater degree.

Defense organizations already have innumerable sources of data—enterprise data, transaction data and real-time detailed data such as weather information and sensor feeds. The information sources for making the best decisions exist right now. Likewise, in terms of processing power and speed, the capabilities of today’s systems mean that even vast amounts of data can be manipulated in real time.

Incorporating predictive insight can allow military commanders to harness the power of their existing infrastructure to analyze this and other data to predict events with incredible insight. And while predictive insight might not give definitive answers, it certainly could tip the scales in the user’s favor. Having the capability to model realistic scenarios with a multitude of variables—in the moment—means being able far better to anticipate an enemy’s movement and the potential implications of every action. Commanders may even uncover possible outcomes that they had not considered before.

A predictive insight loop can be considered the next generation of the military’s familiar observe, orient, decide and act (OODA) loop, using analytics to execute faster. The process begins by pulling together physical world observations, such as those made possible through radio frequency identification (RFID) technology and biological sensors, as well as data for each military asset, including vehicles and weapons systems components. While this information typically is available within the operation as a whole, rarely is it found in a centralized space. It tends to be spread across units and buried in stovepiped systems, which limits its immediate usefulness. Its aggregation is a critical first step.

Once the information is compiled, commanders can begin to orient their response by running micro- and macro-scenarios. For example, models that build on sensory data from numerous component parts can predict how an individual machine is going to perform the next day, at a specific time, on specific terrain in specific weather conditions. However, instead of a user having to make educated guesses based on generic thresholds for a type of asset, the predictive approach provides individualized models for a particular asset.

This is a big difference compared with conventional approaches. Considering how machinery often is manufactured through automation, it would be reasonable to expect that each piece rolling off an assembly line is identical. But machines have their quirks. They develop personal histories based on how they are used over time. A new car given to a newly licensed teenager may be a very different vehicle in just a year’s time compared with the one behind it on the assembly line that was bought by a Sunday driver. Similarly, a troop transport that operates in desert conditions will develop a different profile than one that always operates in a cool climate.

The information that comes from predictive insight helps shape the most appropriate response given a very specific context. This has implications starting with basic expenditures. Predictive insight allows for performing maintenance when it must be done and not on a predefined schedule, thus optimizing the time between routine maintenance.

Another benefit is the ability to minimize offline time from unexpected breakdowns. Users can know when failures are imminent and can order parts and schedule repairs in advance. This concept already has been proved in practice outside the defense realm. Metropolitan St. Louis has used it to enhance fleet management for its city buses to minimize downtime and to drastically cut maintenance costs bus by bus.

Predictive insight also could help performance-based logistics take root. Currently, the total cost of an engine is not its initial cost; it is that cost plus maintenance and repair over the long term. In the near future, rather than buying an engine and then paying the purchase price over again for parts and service during the life of the engine, military units will be able to purchase “power by the hour.” It is paying for performance—or buying uptime. The closest analogy would be civilian cell phone call plans with set minutes. For operating hardware, the military would pay for a minimum service level. When it wanted more, it would pay more.

With this approach, the military can eliminate the risks associated with maintaining an aging fleet. And, the concept gives a new incentive for vendors to reach new levels of product excellence. Because vendors will want to maintain a lean inventory, they will need to know what is on a particular piece of equipment at any given time and to be able to predict when failure will happen. They will build predictive insight capabilities into their machinery right from the start.

 

Controllers in the air traffic control center aboard the USS George Washington monitor the launch and recovery of aircraft. Aspects of military operations ranging from tactical decisions to logistical support and maintenance can be improved significantly by employing predictive insight.
©Reuters/Corbis

This model presumes closer partnerships with the private sector and a whole new level of data sharing. Vendors will need more information to support battle plans. For example, if a vendor manages to achieve a 100-hour-per-month service level for three consecutive months—but will need to provide 120 hours for training exercises the next month—that vendor will need to be able to prepare to provide that capacity. The military purchaser will need to communicate information into the picture so that the predictive models will work and, ultimately, keep costs down for both parties.

Beyond equipment and machinery at the operational level, the power of predictive insight can assist decision-makers in mission planning by predicting the optimal mix of skills for different scenarios, adapting to the probable reactions of an enemy or even reviewing alternative targeting runs. Military commanders can decide how best to act when they have a broader range of potential scenarios and corresponding options for action in front of them.

Having a predictive insight system raise red flags will allow commanders to resolve issues before they become obvious. The last thing any military leader wants is for a problem to manifest itself in the course of heavy action, when the stakes are highest and failure is most costly. For example, it is far preferable to pull a troop transport aside proactively than to let it break down in the middle of a road. But under a traditional maintenance model, all too often a mechanic might not have any inkling that a problem even exists. Without predictive insight, taking pre-emptive action could be a costly and pointless shot in the dark.

The potential for predictive insight spreads beyond the mechanical to more strategic, macro decisions—helping commanders plan deployments, for example, or use RFID and global positioning system technology to determine optimal logistics operations and to prevent and respond to breaks in the supply chain.

For the military to be able to harness the power of predictive insight fully, it not only must have the information available, it also needs the ability to act on it. The first half of this equation already is in place. Making predictive insight possible is not going to require huge outlays of capital on infrastructure. The technology that exists now already has the capabilities to collect the necessary information or can be retrofit at a low cost. There is no lack of data available. The data needed to put predictive insight into practice exists and probably already is being collected in some fashion. The military need only decide that it is worth taking the time to pull it all together.

What still is needed is the ability to analyze the data and present the analysis in a usable fashion so that it becomes actionable information. Most importantly, an organization must have the creativity and flexibility to let go of convention so that when predictive insight makes the optimal response clear, those with the authority to act are not hamstrung by old ways of doing things.

High-performance warfighters out-execute their foes. Predictive insight offers an invaluable advantage in this regard: the opportunity to stay one step ahead not only of the enemy but also of the current reality. Ultimately, the predictive insight approach could transform the face of warfare by building a 360-degree forecast system for command and control operations.

 

Lt. Gen. Jack Woodward, USAF (Ret.), is a senior executive at Accenture. Ryan M. LaSalle is an information technology consolidation and advanced technology expert at Accenture Technology Labs.