Autonomous Combat

UAVs as an Instrument of War

2009-11-10T09:25:00.000-08:00

The National Defense University is holding a conference on the use of UAVs in war. The conference is December 14-15, 2009 at Fort McNair, D.C. Registration is on-line at: http://www.blogger.com/www.ndu.edu/CTNSP/Event_Registration/register.cfm. There are several interesting topics including:

UAV Roles and Missions
Theme: The number of UAVs and their missions are increasing rapidly. Where should the line be drawn? Is it necessary to have so many different systems and who should operate them?

Modern UAVs: Designs, Payloads, and Capabilities
Theme: While drone platforms are being created to carry bigger ordnance, there are other designs for unmanned fighter planes, robotic flying ICUs, and bio-inspired nano-autonomous systems. How far will this technology go? Will platforms become more autonomous or are there certain capabilities that need to maintain a human face?

Rise of the Machines: The Autonomous Line in the Sand
Theme: With some scientists talking about the possibility of creating super soldiers (“Iron Men”) and others developing adaptive artificial intelligences, the melding of men and machines seems almost inevitable. But should it be? Should the fact that the humans who develop these systems make mistakes give us pause? If we create computer systems that have greater computing power than the human brain, is it possible that we could be replaced? Where is the autonomous line in the sand that we shouldn’t cross?

The HALO Postulate: Is War Becoming A Game?
Is the prevalence of games such as HALO and Call of Duty blurring the lines between reality and simulation for soldiers? With fewer and fewer politicians that have experienced combat, is technology making it easier to use the military option first instead of last?

The Ethical and Legal Implications of Using UAVs
Theme: Where do UAVs fit in the overall American arsenal? Do we have a strategy for integrating these drones into existing protocols in the United States and abroad? Can we really expect robots to tell the difference between combatants and non-combatants when our own soldiers have difficulty doing so? Do these platforms conflict with the Geneva Conventions and the International Committee of the Red Cross’s regulations for weapons of war?

Artificial Intelligence

2009-08-31T18:48:00.000-07:00

True autonomous systems will require some level of decision making ability. Will this be a complex expert system or will it truly be artificial intelligence (AI)? I first studied AI nearly 20 years ago. Back then, there was great hope and excitement for the possibility of true AI in the not so distant future. And then nothing happened. Computers are immensely more powerful today yet they remain unable to match the thought and creativity of a child. I believe that autonomous systems of the next decade will be highly complex yet capable expert systems. We will only trust them to operate within pre-defined envelopes. Our challenge will be to define these performance envelopes in such a way as to enable the systems to address unforeseen circumstances within their envelopes.

Halal [1] believes that developments in AI will lead us to an “age of consciousness” around 2020-2030. In his vision, AI will free us from thinking about, “routine mental tasks,” so that we may focus on spiritual dilemmas (i.e., competing values, beliefs, etc.). I have two issues with this vision. First, I don’t believe AI will progress as rapidly or effectively as Halal. Second, I do not agree with his assumption that people, freed of menial tasks, will become altruistic and give up their personal beliefs in order to seek the common good.

[1] Halal, William. (2008). Technology’s promise. Expert knowledge on the transformation of business and society. Palgrave Macmillan. New York.

Predicting the Future

2009-08-29T20:22:00.000-07:00

William Halal presents many thought provoking ideas in his book Technology’s Promise [1]. As a pilot, his predictions regarding small aircraft are of special interest to me. Halal predicts that small aircraft will account for 30% of all air travel by 2025. I find this difficult to believe. Halal supports his position with statistics such as 98% of Americans living within 30 minutes of a small airport, the $150,000 price of some small aircraft, and the number of “micro-jets” under development. However, I think his assessment of the facts is, shall we say, overly optimistic. First, the thirty minute drive to the small airport may be 30 minutes in the wrong direction. Furthermore, getting to the airport is just the beginning. The aircraft must undergo a safety inspection before departing and then the traveler must wait his turn to depart the airport. These factors can easily add an hour from the time you arrive at the airport. As for the price of aircraft, the $150K he mentions is typically not for an aircraft capable of flying in instrument conditions (i.e., clouds and bad weather). Even if it were, what percentage of the population can afford that expense, not to mention the thousands of dollars required for annual maintenance and inspections? The economics of aviation are daunting. In fact, several of the exemplars Halal mentioned have already gone bankrupt (not to mention their target price points are usually over $1,000,000 per plane). Additionally, his prediction of small aircraft accounting for 30% of travel is problematic. From June 2008 to May 2009, 718,389,000 domestic passengers traveled on commercial flights [2]. Even assuming the average passenger was traveling in a group of 2, shifting this traffic to small aircraft (which typically can only carry 2 passengers) would put an additional 107,758,350 flights in the air annually. That is roughly ten times the current number of commercial flights. Halal’s belief that technology will overcome this is admirable but misplaced. Even if the technology is available, it is unlikely to be in place. The average age of small aircraft in the U.S. is approximately 30 years. Some estimates show this will approach 50 years by 2020 [3]. The majority of these aircraft still uses analog instruments and has yet to have “glass cockpits”, GPS, or other technology of the past decade installed. Based on this track record, it seems unrealistic to believe they would receive the technology necessary for “free flight” in the next decade.

Flying is inherently more expensive and more risky than traveling by automobile. Bad judgment tends to be fatal in the air where you cannot simply pull over to the side of the road. Think of all the people you see every day run red lights or cause accidents because they are talking on the cell phone. Now imagine that when something goes wrong they fall 10,000 feet to the ground. All in all, Halal’s vision, while intriguing, seems misguided, uninformed, and overly optimistic to the point of simply being wishful thinking. It feels a lot like a revised vision of the personal jetpack that also failed to materialize.

[1] Halal, William. (2008). Technology’s promise. Expert knowledge on the transformation of business and society. Palgrave Macmillan. New York.
[2] http://www.bts.gov/xml/air_traffic/src/index.xml#CustomizeTable
[3] http://www.popularaviation.com/docs/agingbestpractices9021.pdf

Global Collaboration

2009-08-28T19:34:00.000-07:00

The Defense Information Systems Agency (DISA) is fielding collaborative tools (a.k.a., Web 2.0 tools). These tools build on an underlying service architecture and communications infrastructureto facilitate knowledge generation and transfer across the DoD.

This collaboration service enables synchronous and asynchronous communication using instant messaging, low-bandwidth text chat, and web conferencing. Instant messaging and web conferencing both include text-based communication, while web conferencing adds shared whiteboards, desktop & application sharing, and the ability to invite non-DoD personnel into collaboration sessions. The collaboration tool is called Defense Connect Online (DCO). DCO is based on Adobe Acrobat Connect Professional and Jabber MomentIM. Collaborative tools such as these have been employed in Iraq and Afghanistan to enable troops to share, in real-time, tactics and techniques they have developed under fire.

Capturing the Future

2009-08-09T20:16:00.000-07:00

Christakis introduces the concept of Structured Design Process (SDP) [1]. In the case study "Anticipating Alternative Futures in Energy Efficiency", Christakis applies the SDP to collaboratively brainstorm, with stakeholders, possible influential factors for promoting energy efficiency in the Pacific Northwest. The SDP is then applied to assess the plausibility of the ideas and determine interrelationships between them. Then, collaborative brainstorming produces a set of possible options for promoting these factors. Finally, individual and group voting reduces the set of options and produces a Consensus Regional Future.

This case study essentially tells a story of consensus building. There is potential in the SDP to leverage the power of collective knowledge. More knowledgeable team members can guide the less knowledgeable. However, there is also potential that the process simply captures common beliefs or the politically correct values that members feel they should promote. Anyone implementing the SDP should take measures to avoid "group think".

[1] Christakis, A.N. and Bausch, K. (2006). How People Harness Their Collective Wisdom and Power. Information Age Publishing, Connecticut.

The Information Age

2009-08-02T11:41:00.000-07:00

Alvin Toffler, in The Third Wave, introduced the concept of waves of change that fundamentally alter the very nature of civilization and destroy the preceding civilization. He asserts that civilization has moved from an Agricultural Age to an Industrial Age and is presently moving into an Information Age. In this Information Age, the nature of every aspect of civilization, including economy and family structure, will be fundamentally and irreversibly changed by widespread access to information.

There has been widespread adoption of the term Information Age. However, it is not clear that the type of fundamental change to civilization Toffler described is actually occurring. Information has not become the basis of the economy as he predicted. Likewise, social structure has changed less due to information than it did as a result of industrial age phenomenon such factories and rapid transportation which promoted the breakup of the traditional nuclear family and changed social values. In contrast, increasing access to information appears to be making largely incremental changes in social behavior and economics. The majority of working Americans still physically travel to their place of employment and, whether conducted in a “virtual” store or a “brick and mortar” store, commerce remains fundamentally unchanged; physical products are produced, marketed, and sold.

There have been claims that the nature of conflict in the Information Age will be fundamentally different from conflict in previous ages. It seems this claim may be somewhat overstated. Certainly, technology has enabled a much faster pace for military operations and has made weapons far more accurate and effective than at any previous point in history. However, precision weapons and unmanned vehicles are simply refinements of existing capabilities. Even systems which collate and process vast amounts of intelligence data are nothing more than improvements on age old processes. Some would argue that the emerging field of Cyber Warfare is the entirely new nature of conflict Toffler predicted. However, it seems to me the strongest claim that can be made is that this is a new domain wherein the traditional concept of conflict will be conducted.

Humans are becoming increasingly effective at conflict. However, the fundamental nature of conflict does not appear to be changing dramatically. The specific enabling technologies undeniably are changing but the basic objectives and types of capabilities are not.

War 2.0

2009-07-25T14:23:00.000-07:00

Web 2.0 has become a bit of a buzz-word. However, researchers and developers are creating interactive programs that pull content from multiple data sources to make synergistic applications that are more useful than the sum of their parts. The MITRE corporation is researching this area with its Tagged and Geotemporal Reporting (TAGR) project. TAGR uses the IBM Mashup Framework to provide situational awareness and redirection of unmanned devices. This type of composable capability could serve as a data source for future autonomous systems. However, there may be implications that warrant further consideration. For example, how can the accuracy and timeliness of mashed up information be established and how do you reconcile conflicting sources of information? Humans deal with these types of issues daily. However, it would seem that autonomous systems would need to be given similar decision making abilities.

Technology on the Horizon

2009-07-25T10:30:00.000-07:00

The Horizon Report mentions a technology that I believe will be relevant for autonomous combat systems, Semantic-Aware Applications. Such applications would use the meaning of information to make connections between information in order to provide contextually relevant answers. This, combined with other technology such as knowledge management and data mining tools, would facilitate autonomous decision making systems. The Horizon Report puts this technology 4-5 years down the road. This coincides well with the timeline for unmanned combat systems evolution. I do not believe the capability I envision will be mature in that time frame. However, I think it is reasonable to anticipate early implementations of semantic-aware applications to support these systems in that time frame.

The Horizon Report also identifies the concept of "collective intelligence" as a key trend as we move forward. The idea of this concept is loosely that a group should be able to reach a better decision than an individual. MIT has created the Center for Collective Intelligence to explore, "how can people and computers be connected so that—collectively—they act more intelligently than any individuals, groups, or computers have ever done before". This is basically the idea that collaboration will result in better performance than individual actors. This concept is a basic tenet of the DoD theory of net-centricity. Therefore, such research should guide the future evolution of the DoD enterprise.

Robots of War

2009-07-14T00:01:00.000-07:00

PW Singer presents an interesting perspective on the impacts robots are having on human warfare in this video clip http://www.ted.com/talks/lang/eng/pw_singer_on_robots_of_war.html

Will autonomous combat systems work in favor of large nation states or smaller non-state actors?

Will the use of autonomous combat systems make nations more likely to go to war?

Opening thoughts on autonomy and power

2009-07-13T22:36:00.000-07:00

The amount of data exchanged between combat systems is rapidly increasing. This presents a dilemma for battlefield systems where the lack of infrastructure forces reliance on radio communications. There is a finite amount of radio spectrum and the amount of information that can be communicated in this spectrum is limited by physics. Furthermore, radio signals take time to propogate and this can lead to latencies of several seconds which can be catastrophic in a combat environment.

Battlefield systems will become increasingly autonomous. The systems will execute combat maneuvers (e.g. dogfight) and engage targets (guns, missiles, bombs, etc.) without direct human control. This will allow systems such as unmanned fighters and bombers to overcome the limitations of radio communications. The systems will be able to identify enemy forces and select appropriate offensive and defensive actions without human intervention. The only thing a human operator will need to do is assign an objective or mission. This will enable a single remote operator to command groups of vehicles.

Autonomous systems will be required to operate under human supervision, as opposed to control. However, that supervision may not be continuous. This is the same way many human teams work. The supervisor/commander assigns the target/mission and dispatches a team to carry out the task. It is common for teams to operate without continuous supervision. However, before people are allowed to operate independently in this manner, they are typically required to demonstrate a specific set of knowledge and skills. Likewise, an autonomous system would have to demonstrate its ability to operate properly across its full operational envelope. We'll never be 100% confident that they won't make an error. However, the same is true of humans. We'll need a review process to assess and address the root cause of autonomous system errors just like we do for human errors.

Another issue for combat forces is power. This has been a major issue since World War II where German tanks had to wait for horse drawn fuel wagons to catch up. Today, we must ship tons of fuel to power engines and generators whenever and wherever we deploy forces. Fuel also limits the flight time of unmanned vehicles. Alternative energy sources will be developed and integrated into combat systems. This will mitigate our dependence on transported fuel and enable missions of greater range, speed, and duration. One exciting possibility is photovoltaic paint that would turn nearly any surface into a solar cell.