Logistics pose the most significant challenge to any military operation. It seems like a simple concept until you factor in the enemy, personnel constraints, maintenance issues, and the operational plan. Fortunately, artificial intelligence (AI) can help solve some of our ground transportation problems. So, where does vehicle automation fit into the Army’s modernization priorities?
The 2019 Army Modernization Strategy outlines six priorities, two of which –the Optionally Manned Fighting Vehicle and the Robotic Combat Vehicle –will use AI to navigate, conduct surveillance, and remove IEDs. Despite the frequent focus on the importance of logistics to military operations, the modernization strategy does not prioritize the integration of AI into logistics vehicles. The Army needs to revisit this omission. To truly begin the integration of AI into ground vehicles, the Army should shift its investment from complex future combat platforms to semi-autonomous support vehicles. Focusing on sustainment vehicles will enable the Army to quickly integrate mature technology into existing platforms while attempting to address the ethical and safety concerns associated with autonomous combat vehicles.
Current Platform Development
For simplicity, this article will focus on two modernization projects, the Robotic Combat Vehicle and the Expedient Leader-Follower program. Both projects represent an attempt to automate the functions of a human crew and reduce soldier exposure.
Major General Ross Coffman, director of the Next Generation Combat Vehicle (NGCV) Cross-Functional Team, described the role of the Robotic Combat Vehicle as acting as the “eyes and ears” for armored formations. Officers on the NGCV Cross-Functional Team provided a succinct description of its potential capabilities. According to the team, it will enable commanders to better employ their forces through the early detection of enemy forces that minimizes exposure to soldiers. Robotic Combat Vehicles will also assist in conducting breaches, long-duration operations, and combat in urban environments. In terms of autonomy, the cross-functional team wants the Robotic Combat Vehicle to move “to the probable line of direct fire contact at a relevant speed with minimal human interactions while intelligently using terrain.”
Meanwhile, efforts to integrate automation into support vehicles are modest. Attempts to incorporate AI primarily revolve around leader-follower vehicles. In September 2020, the Army fielded 60 sets of leader-follower equipment for testing on the Palletized Load System. Current plans for the leader-follower system have a manned vehicle leading four unmanned trucks in convoy operations. Through sensor input, the unmanned vehicles mimic the movements and path of the leader. Trucks equipped with the technology can operate day and night, through various weather conditions, and on all drivable surfaces. Leader-follower developers aim to increase the number of follower vehicles to nine by the 2030s.
Each project receives vastly different levels of funding. In FY 2022, Congress authorized $75 million for the research and development of the Robotic Combat Vehicle, while the leader-follower program received a modest $7.6 million for vehicle testing. Current funding levels provide a glimpse at discrepancies in fielding costs between the two systems. Leaders will have to grapple with which projects to continue as the budget remains relatively flat and research, development, and acquisition funding decreased from $37 billion to $35 billion.
The Limitations of Autonomous Combat Vehicles
Integrating AI into combat platforms presents numerous limitations, but autonomous navigation is one of the most challenging. Off-road navigation presents many challenges that do not exist in urban environments and make its implementation difficult. Like future upgrades to the Robotic Combat Vehicle, autonomous ground vehicles will have to incorporate advanced algorithms that can quickly make decisions in a constantly changing environment. AI will have to respond to changes in the mission, adjust to enemy movements, and adapt to unmapped or unfamiliar terrain instantaneously.
Army researchers will continue to make impressive strides to improve the performance of AI, but the complexity of the decisions suggests that Army may need an artificial general intelligence system. General intelligence is a system “that can perform tasks across domains as skillfully and flexibly as humans.” Autonomous and unmanned systems will have to accomplish the roles of three to four humans, not just one. Unfortunately, a system capable of achieving these tasks may not be available until 2060.
Performing these functions in an autonomous or remote-controlled manner also assumes the platform will not experience hacking, jamming, or signal interruption. Ground vehicles frequently lose communication because of varying terrain and infrastructure, but this problem worsens when you factor in electronic warfare. Addressing this vulnerability is vital for its future implementation.
Though the Robotic Combat Vehicle is not a lethal autonomous weapons system, it presents a potential ethical dilemma. Efforts to ban lethal autonomous weapons have thus far been unsuccessful, but the United Nations continues to try to address the issue. Several of America’s allies and partners strongly resist creating autonomous weapons. Opposition to these systems may hinder their future implementation and forward deployment in Europe and the Indo-Pacific.
Autonomous weapons also pose concerns regarding accountability. Ethics scholars assert that if autonomous weapons engage civilians, it will be difficult to attribute responsibility for the act. Autonomous combat vehicles will require significant levels of trust from ground commanders, and accidents can erode that confidence. Soldiers are typically slow to trust new technology and are often skeptical about its uses.
The United States can examine Russia’s testing of the Uran-9 to guide future autonomous vehicle development. Russia’s Uran-9 is robotically controlled and can reportedly operate four kilometers from its control vehicle. It possesses modest automated functions, including pre-programmed paths and obstacle avoidance. During testing in Syria, the Uran-9 suffered multiple failures in “transportation, communication, firing and the operator’s situational awareness capabilities.” The Uran-9’s malfunctions in Syria pushed many Russian military analysts to suggest unmanned platforms maintain only a limited capacity to conduct low-intensity conflict and reconnaissance operations, suggesting that the technology is not mature enough to field an autonomous combat vehicle.
Prioritize the Automation of Sustainment Platforms
As the Army continues to develop and mature the technology necessary for the Robotic Combat Vehicle, it should increase its investment in autonomous systems that it can implement immediately. Software, such as that used for the leader-follower program, can provide a valuable force-multiplier while generating data and feedback necessary for future autonomous platforms.
Leader-follower technology is currently mature and can drastically reduce the stress of convoy operations on sustainment units. RAND’s 2020 study on automating Army convoys recommended that the Army adopt a minimally manned employment concept for automated trucks. The concept would have only the driver remain in the follower vehicles to oversee the system. RAND found that it would reduce the number of soldiers required for convoy operations by 28 percent and increase soldier productivity by 38 percent.
Extensive training exercises such as those at the National Training Center constantly stress the ability of units to conduct sustainment operations. One of the biggest roadblocks to success is personnel limitations. Sometimes this is due to shortages in personnel, but it’s frequently due to crew exhaustion. Support units attempt to sustain continuous operations while implementing an effective rest plan. Ultimately, an inability to constantly move the classes of supply forward slows the operational tempo. Leader-follower technology may not solve every problem, but it will reduce the stress on sustainers.
Leader-follower technology also alleviates the tension associated with making room in the budget. Since the program does not require the procurement of new vehicles, the Army can distribute an emerging technology without the cost of buying a new vehicle fleet. Army leaders can also export leader-follower software to other platforms and plan to begin the process this year. By taking incremental steps to integrate autonomous software into existing platforms, the Army can potentially reduce the cost of future platform procurement and gain valuable feedback.
The Future of Automation
As the Army continues to integrate autonomous software into its ground vehicles, it should consider focusing its initial efforts on sustainment platforms. Integrating autonomous software into combat vehicles presents many unique challenges that will take time to address. Performance by Russian unmanned systems demonstrate the limitations on current technology.
The 2019 modernization strategy focused heavily on the automation of combat platforms, but the ability to create autonomous support platforms may be more practical. Programs like leader-follower and future autonomous support vehicles bypass the ethical considerations of weapons systems and mitigate the complexity associated with combat vehicles.
Automating the functions of three to four crew members is a difficult task. Future initiatives may look to automate specific tasks inside armored vehicles, like the gunner. AI could increase system reaction time and reduce human error when engaging targets. Incrementally applying AI to certain crew functions will provide new capabilities while we await the creation of a general intelligence AI. Leader-follower does precisely this. It replicates the actions of a human driver and reduces the personnel required for convoys. Getting this software into the hands of soldiers will drive future innovation and improve the quality of future ground vehicles.