Abstract
Robots and their uses are a staple of science fiction, and yet practical applications are already in the field. Uninhabited aerial vehicles, such as Predator or Global Hawk, can be considered robots. This article examines robots by using the battlespace operating systems (BOS) paradigm. The author concludes that, as technological change gathers pace, capability planners need to consider the uses and benefits that robots offer the military of tomorrow.
Introduction
Anyone who has seen the latest episode of Star Wars will have entered a fantasy world populated by ‘robots’ with a versatility that many of us could only have imagined. But just how fantastic is such a vision?
The use of robot technology is not new. During World War II, the German Army employed a tele-operated tracked vehicle known as ‘Goliath’, which was packed with explosives and detonated under advancing Allied tanks. Outside defence, several other domains—including the manufacturing, construction and mining industries—have been successfully using robotics applications for many years. The drivers for such action within industry include the need to remove people from danger, to eradicate boring and repetitive work, and to reduce manpower costs. Many of these driving factors are equally relevant—if not more so—to defence. Robotics is a field that the military can no longer afford to ignore. Capability developers should continue to enjoy the movies but also start to think seriously about using robots for military operations.
My intention in examining this issue is to identify potential applications for robotics in the conduct of land warfare, specifically within each of the battlespace operating systems (BOS).1 I also intend to examine the broad performance parameters for military robots, and touch on some of the issues relating to the use of robotics in the conduct of land warfare.
What is a Robot?
There is no universally agreed definition of what constitutes a robot. For the purposes of this discussion, a robot will be defined as a mechanical device that can be programmed to perform tasks or functions involving movement and manipulation previously performed by humans.2 All robotics systems contain some form of decision-making function—whether it be at the commencement of a mission or during the mission itself. Robotics systems comprise the robots themselves, their operators, and their maintainers. A robotics system might also include a communications system that operates between the robots and their home base or between groups of robots. ‘Robotics’ refers to the application of robots to specific tasks or functions.
Robots may be remotely operated (tele-operated), act semi-autonomously or autonomously. Tele-operated robots generally rely on the operator to sense3 the operating environment and control the performance of the robot in that environment. Semi-autonomous robots—once deployed—are capable of performing some of their intended functions without human intervention. Autonomous robots—again, once deployed—are capable of performing all of their intended functions without human intervention.
Tele-operated and semi-autonomous robotics systems must therefore include a control and communications element so that there can be interaction between the robot and its operator. Semi-autonomy is likely to allow an operator to manage fleets of robots. Additionally, communication between robots will enable them to cooperate and perhaps increase their degree of autonomy. All robotics systems also require materiel and technical support—even if only in the early stages of a task.
In fields other than industrial automation, most robotics applications need to move through the environment and thus the term ‘uninhabited vehicle’ has become fashionable. Uninhabited vehicles include uninhabited aerial vehicles (UAVs), uninhabited ground vehicles (UGVs) and uninhabited underwater vehicles (UUVs). These terms have limited utility as they describe only the environment in which the robot operates and not the function of the robot itself.4 UAVs with a surveillance and reconnaissance function are now evolving to include the ability to fire a weapons system—necessitating a change in term to Uninhabited Combat Aerial Vehicle (UCAV). Given the increasing number of robotics applications, these terms have the potential to become too broad to remain meaningful. A return to the function or task—reconnaissance, for example—is likely.
The Utility of Robotics
The use of robotics offers multiple advantages to military forces in the conduct of land warfare. Like industry, the paramount advantage lies in removing soldiers from immediate danger. The ability to reduce costs (especially manpower) to achieve the same or greater effects is a second persuasive advantage. These advantages alone warrant the evaluation of robotics for application in the conduct of land warfare. Surpassing both these, however, is the fact that robots have the potential to change the way we conduct military operations.
Land Warfare Applications
The single greatest impediment to the development of robotics applications for land warfare at this time is not the limits of technology—it is the lack of guidance from the military on the way in which robotics might be useful in the military context.5
There is one key delineation between military applications, and it concerns the distinction between those that are designed to kill the enemy and those that are not. This delineation is crucial because our present ‘rules’ for warfighting allow the enemy to surrender at the last safe moment before imminent death or injury. Any active engagement of an enemy target by our weapons systems is routinely preceded by a number of checks designed to ensure that the target is legitimate. An ethical issue that arises from this scenario concerns the question of whether it is acceptable to ‘set and forget’ a robot to go forth and kill the enemy, perhaps several days or months hence. Should robots have the means to discriminate—to capture a surrendering adversary rather than simply killing all designated ‘enemy’? Could or should robots be programmed to abide by the applicable rules of engagement? While these issues occupy a future horizon at this stage, the non-lethal applications, on the other hand, have the greatest potential for utility in the near term and are also likely to provide the most return on investment in the immediate future.
I intend now to focus on identifying potential applications for robotics in land warfare, employing the BOS construct as a means to examine these.
The Command, Control and Communications (C3) BOS
The C3 BOS has broad applications across all the other BOS as it is a necessary function for the conduct of warfare. As a result, the application of robotics to the C3 BOS may be relevant in whole or part to each of the other BOS. There are two primary applications for robotics within the C3 BOS. The first of these involves the use of robots as communication re-transmission or re-broadcasting nodes that position and re-position as the force moves, in order to optimise connectivity. The second role for robotics lies in the reduction of deliberate electromagnetic emission signatures through distancing or dispersion.
The Manoeuvre BOS
The specialised requirements of the manoeuvre BOS also give rise to potential robotics applications. These include the remote destruction or neutralisation of the enemy, particularly in conditions hazardous to humans. Likewise, the clearance of mines, booby-traps and other obstacles ahead of friendly movement is a valuable application for robotics. The provision of sentry systems that identify and warn of approaching threats is a role that could be readily adapted to the use of robotics. Equipping these robots to engage an adversary may also be possible once issues surrounding the rules of engagement have been resolved.
Specifically designed robots could locate and attack designated targets including vehicles (for example, tanks), equipment (helicopters and UAV), stores (fuel dumps) and facilities (C2 nodes). The final target on this list is the adversary himself. Imagine the psychological impact on a commander who learns that the enemy has just released 10 000 multi-environment ‘assassination robots’ programmed to locate, identify and assassinate him. Each of these targets has a unique signature. Robots such as these may be the next generation of precision munitions—in fact, it is in this area that robotics has the single greatest potential for changing the way we fight.
In a direct lift from a sci-fi novel, wearable robotic suits could be designed to enhance the performance of individual soldiers. Robotic suits could potentially incorporate climate control, protection, information systems, and first aid or drug regulation.
Robots could also be used to mount decoy manoeuvre BOS and other ‘signature’ vehicles and equipment which move in accordance with a deception plan.
The Intelligence, Surveillance and Reconnaissance (ISR) BOS
The ISR BOS is also applicable across the broad spectrum of other BOS as it involves the acquisition of information about the environment in which forces operate. Thus, the use of robots within the ISR BOS will apply also in whole or part to each of the other BOS. By far the most significant opportunity for the application of robotics to the ISR BOS lies in the conduct of remote reconnaissance and surveillance. This particular application has been well supported with the use of a range of UAVs in recent years. UAVs have been recognised as high performers in this role principally because of their ability to move through an environment that contains relatively few obstacles. Movement along the ground is more challenging, but there remains enormous utility for reconnaissance UGVs and UUVs—particularly in confined, complex and hazardous environments. The environmental boundary is an artificial one and there is tremendous potential for the development of a robot that can fly or swim to a site, conduct its reconnaissance on the ground and then return to the sea or air in order to relay its collected data.
The Information Operations (IO) BOS
The IO BOS involves a number of potential applications for robotics. These include support to deception plans with decoys; support for the conduct of psychological operations; and support for the conduct of electronic warfare and navigation warfare.
The Offensive Support (OS) BOS
The OS BOS is rich in possibilities for the application of robotics, particularly in the area of target acquisition. These applications primarily involve the detection, recognition, identification, location and marking of targets. Robots could also be used in the mounting of decoy high-value targets to unmask enemy weapons systems. Robots could play a role as platforms for the deployment and operation of delivery systems. The mounting of decoy OS BOS vehicles and equipment that move and emit signatures in accordance with a deception plan constitutes a further significant role for robots.
The Mobility and Survivability (M&S) BOS
Potential robotics applications within the M&S BOS include the vital aspect of mobility support. The clearance of land-mines is an area that has already been populated with tele-operated flails and rollers. Clearance of booby-traps, mines and unidentified explosive objects (UXO) through spoofing or by detection and neutralisation provides another role for robotics. Likewise, the reduction of complex obstacles, including the clearance of simple obstacles such as rubble and surface-laid UXO, or cluster munitions from airfields or key areas, presents a further opportunity for the use of robotics. Robots could also be used in the detection, identification and marking of areas contaminated with CBRN and other hazardous materials.
Robots are ideally suited for use in counter-mobility support, particularly in the positioning of demolition charges. They have the potential to be highly effective in the demolition of buildings and bunkers with mass blast effects, although they are probably unsuited for the demolition of bridges or other complex structures. Robotics is already applied through mines and networks of mines that move (independently or on command) and which detonate autonomously (these are currently recognised as ‘fourth generation mines’). Robots could be used also in the construction of complex obstacles including the laying of minefields and digging of anti-tank ditches.
Robots have significant potential in the area of survivability support, particularly in the mounting of decoy vehicles, equipment and multi-spectral smoke generators in accordance with a deception plan. They could be used to mount decoy high-value targets to encourage enemy weapons systems to unmask.
Large-scale earthmoving and construction projects comprise another area in which there is opportunity for the application of robotics. Additionally, robots could be effectively employed to acquire geospatial data and position differential GPS stations.
The Ground-Based Air Defence (GBAD) BOS
Potential robotics applications within the GBAD BOS include target acquisition, particularly in ground-based or airborne threat detection systems. The identification, location, classification and marking of targets are tasks that could also be accomplished by robots. Robots could have a wider application in the deployment and operation of delivery platforms including cooperative systems that optimise coverage and that are networked with other air defence assets. The mounting of decoy GBAD BOS platforms and equipment which move (and emit) in accordance with a deception plan could also be achieved through the clever use of robotics. Assessment of battle damage to confirm the effectiveness of an engagement is yet another task well suited to robots.
The Combat Service Support (CSS) BOS
The CSS BOS is closely linked to all other BOS given the primary need for each to be sustained in order to perform its role. Thus, potential applications for the C3 BOS may be applicable in whole or part to each of the other BOS. In the CSS BOS area of supply, there is potential for the use of robots as follower ‘mules’ for the carriage of stores, weapons and other equipment behind mounted or dismounted BOS elements. Robots could also be used to transport all classes of supply to their users. Leader or follower ‘mule’ combinations in the first generation may be employed to transport supplies in any environment—land, sea or air—and in time may be able to move between environments. Robots may also be utilised as labour-saving devices in the handling of inventories and the packaging of supplies such as bulk water.
Robots could potentially be used to evacuate casualties from forward battle areas to medical facilities. The use of robotics in remote diagnosis and surgery is currently well established in advanced medical centres in the Western world. In terms of equipment repair and maintenance, robots could be used to populate equipment repair facilities and to recover vehicles and equipment, particularly from combat zones.
Robot Performance Parameters
Across the broad range of applications, there appears to be a set of consistent performance parameters for robotics systems against which there is scope for maximum or minimum criteria to be applied. Optimally, maximum criteria would include the ability to operate in a complex warfighting environment; the ability to be integrated with other land force systems; endurance, robustness and survivability; and simplicity and versatility. The minimum criteria applicable would comprise mass and volume; cost; and signature (unless producing a large signature is part of the robot’s function).
Clearly, individual solutions would apply within these parameters. Any robotics system would, however, suggest an immediate improvement to the current approach in terms of performance of function, training cost, supply and maintenance costs.
Issues
There are a number of issues that relate to the use of robotics in the conduct of land warfare. I will touch on those that predominate and look for the opportunity for further discussion of these and other issues in the future. Many of these issues concern the impact of robotics on the rules of engagement as we know them. They raise questions such as:
- Would it be an act of aggression to deploy a robot into another sovereign country’s territory?
- How do we engender human trust in a robot’s performance, particularly for applications such as mine and booby-trap clearance?
- How ‘deep’ into the battlespace would we want to control robots—as far as possible?
- How can we reduce the vulnerability of robotics systems to navigation warfare?
- How do we prevent the capture of any secure communications equipment carried by the robot? Will a robot be able to determine when it has been captured?
- Given that robots are likely to become targets, what degree of protection (physical and electronic) is necessary or cost effective?
- Is there scope for counter-robot robots? What counter-robot measures might an enemy take?
- Could robots conceivably be ‘turned’ against their original operators?
- What counter-robot measures might it be necessary for us to adopt against adversary robotics systems? Should these include physical (fires and obstacles) as well as electronic interdiction?
- How expensive could a robot become before it is judged to be more cost effective to continue using a soldier for the task or function?
The Way Forward
The potential advantages in using robotics in future land warfare demand that robotics applications are, at the very least, considered by capability developers. Clearly, the Army has a moral responsibility to investigate rigorously any action that could reduce the amount of risk to which soldiers are exposed. Indeed, the introduction of robotics applications in the battlespace has the potential to change the type of people the Army seeks to recruit. Contrary to expectations, anemic high school drop-outs with a talent for violent computer games may yet become highly prized assets.
How can we move forward in this area? It is crucial that capability development staff investigate robotic solutions and challenge the traditional paradigms on the achievement of tasks or functions. Some current robotic technologies can be readily brought across from other domains into military applications. Other applications, however, will require more research and development. Certainly the recent agreement between the Defence Science and Technology Organisation and the University of Sydney’s Australian Centre for Field Robotics for the formation of a Centre of Expertise in Defence Autonomous and Uninhabited Vehicles is a step in the right direction. It is vital not to be constrained by environmental boundaries and equally important that the Defence Science and Technology Organisation be well placed to consider issues along technology lines and apply these to the relevant environmental areas. In future, robots may become some of the first truly multi-environment or joint platforms.
Endnotes
1 Readers unfamiliar with the battlespace operating systems (BOS) should refer to Land Warfare Doctrine (LWD)1 The Fundamentals of Land Warfare, Australian Army, 2002, p. 83. The electronic version of this publication can be found at: http://www.defence.gov.au/ARMY/LWD1/LWD1sitemap.htm
2 Given this definition, information technology and computers are not robots in and of themselves although they may be an essential component of robots.
3 This involves sensing either directly or remotely through the robot, albeit not necessarily through purpose-designed sensors.
4 Nor do they describe all the potential environments in which robots are capable of functioning. For example, how should we refer to robots that may travel inside the human body for medical applications?
5 This became clear at the DGLD, DSTO and LWDC-sponsored ‘Robotics for Future Land Warfare’ seminar and workshop, May 23–4, 2002.
