122. The Guy Behind the Guy: AI as the Indispensable Marshal

[Editor’s Note: Mad Scientist Laboratory is pleased to present today’s guest blog post by Mr. Brady Moore and Mr. Chris Sauceda, addressing how Artificial Intelligence (AI) systems and entities conducting machine speed collection, collation, and analysis of battlefield information will free up warfighters and commanders to do what they do best — fight and make decisions, respectively. This Augmented Intelligence will enable commanders to focus on the battle with coup d’œil, or the “stroke of an eye,” maintaining situational awareness on future fights at machine speed, without losing precious time crunching data.]

Jon Favreau’s Mike character (left) is the “guy behind the guy,” to Vince Vaughn’s Trent character (right) in Swingers, directed by Doug Liman, Miramax;(1996) / Source: Pinterest

In the 1996 film Swingers, the characters Trent (played by Vince Vaughn) and Mike (played by Jon Favreau) star as a couple of young guys trying to make it in Hollywood. On a trip to Las Vegas, Trent introduces Mike as “the guy behind the guy” – implying that Mike’s value is that he has the know-how to get things done, acts quickly, and therefore is indispensable to a leading figure. Yes, I’m talking about Artificial Intelligence for Decision-Making on the future battlefield – and “the guy behind the guy” sums up how AI will provide a decisive advantage in Multi-Domain Operations (MDO).

Some of the problems commanders will have on future battlefields will be the same ones they have today and the same ones they had 200 years ago: the friction and fog of war. The rise of information availability and connectivity brings today’s challenges – of which most of us are aware. Advanced adversary technologies will bring future challenges for intelligence gathering, command, communication, mobility, and dispersion. Future commanders and their staffs must be able to deal with both perennial and novel challenges faster than their adversaries, in disadvantageous circumstances we can’t control. “The guy behind the guy” will need to be conversant in vast amounts of information and quick to act.

Louis-Alexandre Berthier was a French Marshal and Vice-Constable of the Empire, and Chief of Staff under Napoleon / oil portrait by Jacques Augustin Catherine Pajou (1766–1828), Source: Wikimedia Commons

In western warfare, the original “guy behind the guy” wasn’t Mike – it was this stunning figure. Marshal Louis-Alexandre Berthier was Napoleon Bonaparte’s Chief of Staff from the start of his first Italian campaign in 1796 until his first abdication in 1814. Famous for rarely sleeping while on campaign, Paul Thiebault said of Berthier in 1796:

“Quite apart from his specialist training as a topographical engineer, he had knowledge and experience of staff work and furthermore a remarkable grasp of everything to do with war. He had also, above all else, the gift of writing a complete order and transmitting it with the utmost speed and clarity…No one could have better suited General Bonaparte, who wanted a man capable of relieving him of all detailed work, to understand him instantly and to foresee what he would need.”

Bonaparte’s military record, his genius for war, and skill as a leader are undisputed, but Berthier so enhanced his capabilities that even Napoleon himself admitted about his absence at Waterloo, “If Berthier had been there, I would not have met this misfortune.”

Augmented Intelligence, where intelligent systems enhance human capabilities (rather than systems that aspire to replicate the full scope of human intelligence), has the potential to act as a digital Chief of Staff to a battlefield commander. Just like Berthier, AI for decision-making would free up leaders to clearly consider more factors and make better decisions – allowing them to command more, and research and analyze less. AI should allow humans to do what they do best in combat – be imaginative, compel others, and act with an inherent intuition, while the AI tool finds, processes, and presents the needed information in time.

So Augmented Intelligence would filter information to prioritize only the most relevant and timely information to help manage today’s information overload, as well as quickly help communicate intent – but what about yesterday’s friction and fog, and tomorrow’s adversary technology? The future battlefield seems like one where U.S. commanders will be starved for the kind of Intelligence, Surveillance, and Reconnaissance (ISR) and communication we are so used to today, a battlefield with contested Electromagnetic Spectrum (EMS) and active cyber effects, whether known or unknown. How can commanders and their staffs begin to overcome challenges we haven’t yet been presented in war?

Average is Over: Powering America Beyond the Age of the Great Stagnation, by Tyler Cowen / Dutton, The Penguin Group, published in 2013

In his 2013 book Average is Over, economist Tyler Cowen examines the way freestyle chess players (who are free to use computers when playing the game) use AI tools to compete and win, and makes some interesting observations that are absolutely applicable to the future of warfare at every level. He finds competitors have to play against foes who have AI tools themselves, and that AI tools make chess move decisions that can be recognized (by people) and countered. The most successful freestyle chess players use a combination of their own knowledge of the game, but pick and choose times and situations to use different kinds of AI throughout a game. Their opponents not only then have to consider which AI is being used against them, but also their human operator’s overall strategy. This combination of Augmented Intelligence with an AI tool, along with natural inclinations and human intuitions will likely result in a powerful equilibrium of human and AI perception, analysis, and ultimately enhanced complex decision-making.

With a well-trained and versatile “guy behind the guy,” a commander and staff could employ different aspects of Augmented Intelligence at different times, based on need or appropriateness. A company commander in a dense urban fight, equipped with an appropriate AI tool – a “guy behind the guy” that helps him make sense of the battlefield – what could that commander accomplish with his company? He could employ the tool to notice things humans don’t – or at least notice them faster and alert him. Changes in historic traffic patterns or electronic signals in an area could indicate an upcoming attack or a fleeing enemy, or the system could let the commander know that just a little more specific data could help establish a pattern where enemy data was scarce. And if the commander was presented with the very complex and large problems that characterize modern dense urban combat, the system could help shrink and sequence problems to make them more solvable – for instance find a good subset of information to experiment with and help prove a hypothesis before trying out a solution in the real world – risking bandwidth instead of blood.

The U.S. strategy for MDO has already identified the critical need to observe, orient, decide, and act faster than our adversaries – multiple AI tools that have all necessary information, and can present it and act quickly will certainly be indispensable to leaders on the battlefield. An AI “guy behind the guy” continuously sizing up the situation, finding the right information and allowing for better, faster decisions in difficult situations is how Augmented Intelligence will best serve leaders in combat and provide battlefield advantage.

If you enjoyed this post, please also read:

… watch Juliane Gallina‘s Arsenal of the Mind presentation at the Mad Scientist Robotics, AI, & Autonomy Visioning Multi Domain Battle in 2030-2050 Conference at Georgia Tech Research Institute, Atlanta, Georgia, on 7-8 March 2017

… and learn more about potential AI battlefield applications in our Crowdsourcing the Future of the AI Battlefield information paper.

Brady Moore is a Senior Enterprise Client Executive at Neudesic in New York City. A graduate of The Citadel, he is a former U.S. Army Infantry and Special Forces officer with service as a leader, planner, and advisor across Iraq, Afghanistan, Africa, and, South Asia. After leaving the Army in 2011, he obtained an MBA at Penn State and worked as an IBM Cognitive Solutions Leader covering analytics, AI, and Machine Learning in National Security. He’s the Junior Vice Commander of VFW Post 2906 in Pompton Lakes, NJ, and Cofounder of the Special Forces Association Chapter 58 in New York City. He also works with Elite Meet as often as he can.

Chris Sauceda is an account manager within the U.S. Army Defense and Intel IBM account, covering Command and Control, Cyber, and Advanced Analytics/ Artificial Intelligence. Chris served on active duty and deployed in support of Operation Iraqi Freedom, and has been in the Defense contracting business for over 13 years. Focused on driving cutting edge technologies to the warfighter, he also currently serves as a Signal Officer in the Texas Military Department.

120. Autonomous Robotic Systems in the Russian Ground Forces

[Editor’s Note: Mad Scientist Laboratory welcomes back returning guest blogger and proclaimed Mad Scientist Mr. Samuel Bendett with today’s post, addressing Russia’s commitment to mass produce independent ground combat robotic systems. Simon Briggs, professor of interdisciplinary arts at the University of Edinburgh, predicts that “in 2030 AI will be in routine use to fight wars and kill people, far more effectively than humans can currently kill.”  Mr. Bendett’s post below addresses the status of current operationally tested and fielded Russian Unmanned Ground Vehicle (UGV) capabilities, and their pivot to acquire systems able to “independently recognize targets, use weapons, and interact in groups and swarms.” (Note:  Some of the embedded links in this post are best accessed using non-DoD networks.)]

Russian Minister of Defense Sergei Shoigu / Source: Wikimedia Commons

Over the past several years, the Russian military has invested heavily in the design, production, and testing of unmanned combat systems. In March 2018, Russian Defense Minister Sergei Shoigu said that mass production of combat robots for the Russian army could begin as early as that year. Now, the Ministry of Defense (MOD) is moving ahead with creating plans for such systems to act independently on the battlefield.

According to the Russian state media (TASS), Russian military robotic complexes (RBCs) will be able to independently recognize targets, use weapons, and interact in groups and swarms. Such plans were stated in the article by the staff of the 3rd Central Scientific Research Institute of the Russian Federation’s MOD.

Uran-6 Airborne Countermine System with flail / Source: Russian Federation MOD

Russia has already tested several Unmanned Ground Vehicles (UGVs) in combat. Its Uran-6, Scarab, and Sphera demining UGVs were rated well by the Russian engineering forces, and there are plans to start acquisition of such vehicles. However, these systems were designed to have their operators close by. When it came to a UGV that was originally built for operator remoteness in potential combat, things got more complicated.

Uran-9 engaging targets with its 30mm 2A72 autocannon on a test range.  Operational tests in Syria proved less successful.  / Source:  YouTube

Russia’s Uran-9 combat UGV experienced a large number of failures when tested in Syria, among them transportation, communication, firing, and situational awareness. The lessons from Uran-9 tests supposedly prompted the Russian military to consider placing more emphasis on using such UGVs as one-off attack vehicles against adversary hard points and stationary targets.

Russian ground combat forces conducting urban operations in Syria / Source: Wikimedia

Nonetheless, the aforementioned TASS article analyzes the general requirements for unmanned military systems employed by Russian ground forces. Among them is the ability to solve tasks in different combat conditions during day and night, under enemy fire, electronic and informational counteraction, in conditions of radiation, chemical contamination, and electromagnetic attack – as well as requirements such as modularity and multifunctionality. The article also points out “the [systems’] ability to independently perform tasks in conditions of ambiguity” – implying the use of Artificial Intelligence.

To achieve these requirements, the creation of an “intelligent decision-making system” is proposed, which will also supervise the use of weapons. “The way out of this situation is the intensification of research on increasing the autonomy of the RBCs and the introduction of intelligent decision-making systems at the control stages, including group, autonomous movement and use of equipment for its intended purpose, including weapons, into military robotics,” the article says.

An example of the complex, ambiguous environments that will challenge future Russian RBCs:  Russian troops in Aleppo, Syria / Source: Wikimedia Commons via article in the University of Melboune’s Pursuit, “Why is Russia Still Supporting Syria?”

The TASS article states that in the near future, the MOD is planning to initiate work aimed at providing technical support for solving this problem set. This research will include domestic laser scanning devices for geographical positioning, the development of methods and equipment for determining the permeability of the soil on which the UGV operates, the development of methods for controlling the military robot in “unstable communications,” and the development of methods for analyzing combat environments such as recognizing scenes, images, and targets.

Successfully employing UGVs in combat requires complicated systems, something that the aforementioned initiatives will seek to address. This work will probably rely on Russia’s Syrian experience, as well as on the current projects and upgrades to Moscow’s growing fleet of combat UGVs. On 24 January 2018, the Kalashnikov Design Bureau that oversees the completion of Uran-9 work admitted that this UGV has been accepted into military service. Although few details were given, the statement did include the fact that this vehicle will be further “refined” based on lessons learned during its Syria deployment, and that the Uran-9 presents “good scientific and technical groundwork for further products.” The extent of upgrades to that vehicle was not given – however, numerous failures in Syrian trials imply that there is lots of work ahead for this project. The statement also indicates that the Uran-9 may be a test-bed for further UGV development, an interesting fact considering the country’s already diverse collection of combat UGVs

As reported in DefenseOne, Russian Colonel Col. Oleg Pomazuev stated that the Nerekhta UGV “outperformed” manned systems in recent exercises / Source: DefenseOne and Sergey Ptichkin / RG

Today, the Russian military is testing and evaluating several systems, such as Nerekhta and Soratnik. The latter was also supposedly tested in “near-combat” conditions, presumably in Syria or elsewhere. The MOD has been testing smaller Platforma-M and large Vikhr combat UGVs, along with other unmanned vehicles. Yet the defining characteristic for these machines so far has been the fact that they were all remote-operated by soldiers, often in near proximity to the machine itself. Endowing these UGVs with more independent decision–making in the “fog of war” via an intelligent command and control system may exponentially increase their combat effectiveness — assuming that such systems can function as planned.

If you enjoyed this post, please also:

Read Mr. Bendett’s previous post, Russian Ground Battlefield Robots: A Candid Evaluation and Ways Forward

… and watch Zvezda Broadcasting‘s video, showing a Vikhr unmanned, tele-operated BMP-3 maneuvering and shooting its 7.62mm MG, 30mm cannon, and automatic grenade launcher on a test range.

Automated lethality is but one of the many Future Operational Environment trends that the U.S. Army’s Mad Scientist Initiative is tracking. Mad Scientist seeks to crowdsource your visions of future combat with our Science Fiction Writing Contest 2019. Our deadline for submission is 1 APRIL 2019, so please review the contest details and associated release form here, get those creative writing juices flowing, and send us your visions of combat in 2030!  Selected submissions may be chosen for publication or a possible future speaking opportunity.

Samuel Bendett is a Researcher at CNA and a Fellow in Russia Studies at the American Foreign Policy Council. He is also a proud Mad Scientist.