Robotic and autonomous systems, deemed necessary for the Army to fight future wars, may be out of its price range unless the service can harness affordably upgradeable, open-architecture systems to keep the cost of acquisition and sustainment down, according to Paul Rogers, director of the Tank Automotive Research, Development and Engineering Center (TARDEC).

“Affordability will kill ground vehicle robotics if we’re not very careful,” Rogers said. “If we are really going to get after this, we have to offer a solution that is a fraction of the cost of a manned combat platform.”

“We have to include in that all of the lifecycle costs,” he added. “Software sustainment is going to be our Achilles’ Heel. I believe software sustainment, over time, will keep the United States Army from fully employing ground vehicle robotics.”

The Army plans on establishing baseline systems with open-architecture software and standard interfaces that will then be continually upgraded over time.

“Proprietary, sole-source solutions are going to be unaffordable going forward,” Rogers said. “That may challenge your business model if you are in industry. So we have to figure out how to merge these two things, because if your business model is not sustainable by the U.S. Army, you’re business is not going to succeed, long-term, anyways.”

Every year the program managers for specific ground combat systems like Abrams or Stryker submit to leadership the office’s need for sustainment funding. Only between 25 and 30 percent of that requirements list is funded, Rogers said. Therefore, program managers are forced to make decisions about what gets fixed and what does not, according to available funding.

“Think about software sustainment for a software-enabled platform,” Rogers said. “If you only get 35 cents on the dollar to sustain it, it is quickly going to atrophy and not be as relevant or reliable when it’s needed.”

For nearly all of its ground vehicles, the Army wants a suite of standard “behaviors” that can be uploaded and performed without a human driver. They include active safety measures like collision avoidance and leader-follower where a single manned truck can lead a convoy of up to seven unmanned vehicles following its exact path.

“We’re working with small vehicles all the way up to combat systems,” like Bradley, Rogers said. Retrofitting legacy platforms with a standard autonomy kit that is upgradable with new software-defined behaviors is a more sustainable acquisition model than the one the military took during the wars in Iraq and Afghanistan.

In the harried search for technologies to detect and defuse roadside bombs, the Defense Department poured funding into robotics programs but ended up with thousands of unique platforms, each with its own controls, interfaces, power supplies and spare parts. Many of those robots have since been divested because their sustainment is unaffordable, Rogers said.

“It was a critical need,” he said. “It was the right thing to do. We spent an incredible amount of treasure on systems. We used them during the war. We came out and had far more than we needed and disposed of thousands of them.”

“That cannot be the standard model going forward,” he added. “Especially when we talk about larger, capital systems, we have to be able to sustain and upgrade very easily over time.”

Establishing a standard, government managed architecture for autonomous systems and software will allow for that upgradeability, Rogers said. Once the baseline hardware is established, then third-party developers can offer new software-defined autonomous behaviors for the range of combat platforms from trucks to tanks, as long as the software is certified to the government interface standards.

“What we want to get down to is this more open, apps-based solution for our capabilities,” he said.