SCHUYLER MOORE
In the summer of 1943, the American submarine USS Tinosa came upon the Tonan Maru, a Japanese oil tanker supporting the imperial fleet. The American submarine spent an entire day and 15 torpedoes trying to sink the Maru, which escaped with light damage after all but one of the U.S. weapons failed to explode on target. Sadly, this episode was hardly unique: more than 70 percent of the sub fleet’s Mark XIV torpedoes malfunctioned during the first two years of World War II.It turned out that the Mark XIV had only been tested in shallow, calm water, at targets that were closer than was realistic, and from barges that did not accurately simulate the effect of firing from submarines. Most importantly, few submariners were part of the testing. When the torpedoes were fired in the deep, choppy water of the Pacific, they neither ran correctly nor consistently detonated.
The enduring lesson? Mature technologies must be put in the hands of the user as quickly as possible to debug, iterate, and improve their capabilities. The commercial sector intuitively knows this. Apple would never release an iPhone without extensive tests and user feedback. (And when it doesn’t, it accidentally releases a phone that doesn’t work for left-handed users; “Antennagate,” anyone?)
Real-world testing is even more important for military equipment. Apple may replicate how and where someone might use an iPhone with some precision, but you will have a harder time precisely replicating how a soldier, sailor, airman, or guardian would use technology in the field, where performance is affected by temperature, geography, weather, and available facilities. A quadcopter might have a perfect record from a company’s launch pad in the U.S. but struggle to take off in the windy mountains of Germany. A fast boat might run well off the coast of Rhode Island, but only half as long in the hot, salty, choppy waters of the Arabian Gulf.
This rule doesn’t only apply to hardware; software must equally be tested in theater. A computer-vision algorithm should be trained with real, recent images from the location it will be used, and a system detecting air threats should be tested against the actual threats it will encounter at a given base. The list goes on.
Military users in the real world are the ultimate gauntlet to determine usefulness of technology. At a remote outpost in Iraq or on a ship in the Pacific, functionality is king; buzzwords and hypotheticals serve no purpose. The technology either works to support the mission, or it doesn’t.
For that reason, the military must push more operational experimentation. While early-stage tech research and development can and should remain at the labs and with the development communities, more mature technology should be pressure-tested in a real environment by real users at the earliest opportunity. The Army’s XVIII Airborne Corps and the Navy’s Task Force 59 are already rapidly driving technology integration out in the real world, but this model of tech experimentation has not yet been accepted and adopted broadly. The military must shift its approach to technology experimentation focused on real user testing in real-world environments if it hopes to keep pace with rapidly-evolving threats and technologies.
U.S. Central Command has already accepted this reality and is moving out. Within the last year, CENTCOM stood up experimentation cells in its Army, Navy, and Air Force Component Commands and is driving technology integration through iterative sprints and exercises in its theater of responsibility. By pushing technologies to their physical limits in a challenging environment with extreme heat, wind, sand, and salt, CENTCOM can serve as the experimentation sandbox for the Defense Department.
It can also test technologies against diverse real-world threats. CENTCOM faces activities ranging from drug smuggling to state-level aggression and conflict, with threats from advanced drones and ballistic missiles at a scale and regularity that cannot be found in other theaters. CENTCOM can iteratively test technologies against increasingly more challenging environments and threats, and is ideally positioned to conduct this experimentation while reducing the risk of catastrophic failure and escalation.
The shift to edge experimentation will require a coordinated push from the policy community and an acceptance of responsibility from the operator community, but it cannot come soon enough. We need to work through the inevitable friction and pain points of technology integration now, or risk repeating the failures that the torpedo community experienced in WWII. The stakes are too high to work through pain points once the shooting starts, and the threat is evolving too quickly to take two-plus years to fix technology failures. The path to rapid technology adoption for the military exists with edge experimentation. We just have to take it.
No comments:
Post a Comment