Bots on the Ground

 

Photo Credit: U.S. Army (per DefenseNews.com)

As a kid growing up in the 90s, I came from a military family. One grandfather served at Pearl Harbor during the bombing (he survived). The other fought in the Korean War. My Dad was OSS during Viet Nam, and both of his brothers were in the Navy- one also in Viet Nam, the other in Operation Desert Storm. My oldest brother joined the Navy when he was old enough, and since I couldn't swim, I went into the Army when I graduated high school. I have to wonder if the military would have been such an obvious and appealing career path if I hadn't had such a deep level of connection to it already. That seems to be the case for potential enlistees nowadays. As the number of veterans decreases, recruiting numbers have plummeted.

In 2018, the military missed its recruitment numbers for the first time ever, coming in at only 7/8ths of their original plan and opting to reduce their future quotas significantly. According to Fortune, "[i]n the 1990s, about 40% of the youth market had parents who served in the military; that number is now down to 15%.". According to Katherine Helland, director of Joint Advertising Market Research & Studies (JAMRS) at the Department of Defense, "[w]e don’t have those intimate connections to military service anymore". So we have less people willing to bolster our ranks. 

What does that have to do with unmanned systems? Unmanned Ground Systems provide the military with the unique opportunity to do more with less.

The United States military has been making efforts for almost a hundred years on various unmanned ground vehicle (UGV) technologies. One would think that with the exponential growth within artificial intelligence and robotics, replacing our warriors with machines would be the next logical step. Given the shrinking numbers of willing human volunteers, it even seems like the inevitable outcome. In their paper, “Bots on the ground: an impending UGV revolution in military affairs?”, author Ash Rossiter disagrees with this assumption, and so do I. 

UGVs have been in use since World War II, in the form of ordnance and weapon-carrying, remote-controlled vehicles called “tele-tanks” (Rossiter, 2020). Since that point, UGV technology has been harnessed for use in surveillance, Logistics support (Rossiter, 2020), mine detection and neutralization (via the Multifunctional Utility/Logistics and Equipment vehicle or “MULE”), and even evacuating injured people from dangerous areas (Ruppert, 2010). There are certainly avenues of potential use. However, there are also technical hurdles which have yet to be cleared.

It seems nearly impossible to develop a UGV that is capable of traversing complex terrain as easily as a human being. AI has not easily found ways of adapting to these challenges. Utilizing remotely operated UGVs in these environments is not much better, as the operators frequently misjudge distances due to the lack of depth perception on their monitor (Rossiter, 2020). 

“BigDog”, the incredible robot developed by Boston Dynamics via their contract with the Defense Advanced Research Projects Agency (DARPA), eventually gave way to “Spot” (see the video embedded below). Boston Dynamics began work on BigDog in 2004. In 17 years, they’ve made massive progress. However, the battery life for Spot is only about 90 minutes. Hardly a game-changer on the battlefield. 

So many of these technologies are created and tested and created and tested, without being fielded in a real-world test bed or given to troops to utilize. The technologies may continue to evolve, but without adequate field testing in comparable settings to real world battle, there’s no way to know if they’ll ever be useful.

There also exists an ethical conundrum with unmanned vehicles that we have begun to experience already with the public’s concern over aerial drone strikes. Either the unmanned system will require an operator anyway, in which case we would still need to enlist people to perform that role, or it will have to be fully autonomous. These UGVs that are capable of killing are called "lethal autonomous weapons systems" (LAWS). There is an understandable level of fear when it comes to giving LAWS the power and intention to take a human life. 

The look of the battlefield is bound to continue changing. As anyone who has served can tell you, though, the gear will forever lag behind the times.

(Video Credit: Boston Dynamics)

References: 

Fortune,  (2020, February 22). Facing falling enlistment numbers, the U.S. Army takes a new approach to recruitment: Mom and Dad.. Fortune. https://fortune.com/2020/02/20/army-military-enlistment-recruitment-ads/

Rossiter, A. (2020). Bots on the ground: an impending UGV revolution in military affairs? Small Wars & Insurgencies, 31(4), 851–873. https://doi.org/10.1080/09592318.2020.1743484

Ruppert, B. (2010, November 22). Robots to rescue wounded on battlefield. Www.Army.Mil. https://www.army.mil/article/48456/robots_to_rescue_wounded_on_battlefield

Risky Business

 In the civilian world, when we think about ships being used outside of the military, three major uses that come to mind are:

1. shipping of goods (e.g. cargo ships or tankers)

2. “roll on, roll off” vehicle transport (a.k.a. ferries)

3. research efforts 

Ferries and maritime research vessels may experience some similar risks as shipping vehicles, but for right now, let us focus on cargo container ships which are used to transport goods over the ocean. Specifically, let us address the recent push to develop unmanned shipping vessels. To clarify, “unmanned” is a term which can have multiple connotations, and can be misinterpreted to mean “fully autonomous”. In this context, unmanned ships will have high levels of automation, such as navigation and obstacle avoidance. They may or may not be remotely operated, and they may or may not have any crew aboard at all. Unmanned here does not mean fully autonomous.

In The Ocean-Going Autonomous Ship—Challenges and Threats, Felski and Zwolak (2020) discuss real world complications of navigating unmanned ships within busy shipping lanes. 90% of all goods (APANEWS, 2019), and nearly 60% of all food (Ritchie, 2020), are transported by ship every year, making many routes quite congested. For an example, look no further than the recent debacle with the Ever Given, a ship which became wedged in a narrow area of the Suez Canal. This one ship crippled a shipping route that is so popular, the loss of its use represented $15 million per day of lost revenue to the country of Egypt, and millions more lost to companies whose livestock perished, foods spoiled, and schedules were upset by the error (Nagourney, 2021). Overseas shipping is big business.

Global Positioning System (GPS) and Global Navigation Satellite Systems (GNSS) are required in order for ships to navigate safely and accurately through the oceans. On a standard system, the same satellite communication array that ships utilize for their GPS and GNSS navigational data are also utilized by the ship’s remote operators to command the vessel. This lack of redundancy is more than a risk of non-redundancy; it can also cause misalignment with the navigational system.

The satellite-tracking functions of the satellite communication terminal use not only raw GNSS and heading data, but augment it with inertial motion units (IMU) and process the positioning information using the algorithms based on Kalman filtering to ensure the proper satellite tracking. Therefore, the lack of GNSS position input will not cause the malfunction of the satellite communication immediately, but the accuracy of the tracking solution will decrease with time. (Felski & Zwolak, 2020, p. 6)

The authors state, and I agree, that the mitigation strategy for this risk should be to duplicate the satellite communication links so that there is both a backup system and a separation between navigation and communications. But there are other threats that exist to the use of GNSS itself.

“Radio frequency interference can be unintentionally emitted by commercial high power transmitters, ultra-wideband radar, television, very-high frequency (VHF), mobile satellite services and personal electronic devices.” (Felski & Zwolak, 2020, p. 6). As the Earth becomes more connected via technology, the presence of potential sources of interference will also increase. On a more sinister note, there have already been documented instances of intentional jamming of GNSS signals against sea-going vessels. The very idea of exploiting unmanned navigational tools is a real threat to the widespread adoption of unmanned shipping.

Here, the authors discuss two potential ways to lessen the impact of GNSS dependence: swarm communication to prevent collision, and pre-planned routes (Felski & Zwolak, 2020, p. 10). In order to combat jamming, though, the author has recommended using 3 separate receivers spread out such that they would be able to identify and compensate for any attempt to jam.

Clearly, the maritime world has a lot to consider when it comes to transitioning ocean shipping to fully automated and remote operated ships. In our next post, I’ll touch on the positives of this move. With all new technology there will be “growing pains”. If we keep looking forward, while learning from what’s behind us, I feel that these issues can be mitigated in a way that makes international maritime shipping safer, better for the environment, and eventually faster.

More later...

(sources)

APANEWS. (2019, June 10). 90 percent of world trade is by sea- Official. Apanews.Net. http://apanews.net/en/news/90-percent-of-world-trade-is-by-sea-official

Felski, A., & Zwolak, K. (2020). The Ocean-Going Autonomous Ship—Challenges and Threats. Journal of Marine Science and Engineering, 8(1), 41. https://doi.org/10.3390/jmse8010041

Nagourney, E. (2021, March 30). With the Suez Canal Unblocked, the World’s Commerce Resumes Its Course. The New York Times. https://www.nytimes.com/live/2021/03/29/world/suez-canal-stuck-ship

Ritchie, H. (2020, January 28). Very little of global food is transported by air; this greatly reduces the climate benefits of eating local. Our World in Data. https://ourworldindata.org/food-transport-by-mode

What the heck are 'unmanned systems'??

    Humans have sought so automate processes for thousands of years. From the mechanization of windmills used to grind grains and water wheels to cut massive pieces of lumber, all the way to fully automated aircraft, we have found better and better ways for machines to do the work of people. As we have moved deeper into the Computer Age, we have begun developing "intelligent" computer systems-- those which can take in data, analyze them, and make decisions independent of a need for human intervention. We are entering into the era of unmanned intelligent technologies.

    But what does that even mean?

    The international community has established criteria to describe levels of automation, so that we can more clearly delineate between features which support operation versus features which work together to provide automation. In my estimation, saying that a system is fully unmanned would require it to be level 4 or level 5 on the SAE J3016 Levels of Driving Automation as listed in figure 1 below (https://www.sae.org/news/press-room/2018/12/sae-international-releases-updated-visual-chart-for-its-%E2%80%9Clevels-of-driving-automation%E2%80%9D-standard-for-self-driving-vehicles). This means that these systems may not function in all conditions, but in the conditions they are designed to support, they will not need any human intervention at all. They aren't just programmed to accomplish a task; they must also be able to sense their environment and make changes to their proscribed actions based upon immediate needs.

    Many people think of automated cars when the word "unmanned" is used, and there are certainly some cars that meet that criteria. However, this also includes simpler things like automated lawnmowers, which can sense obstacles and weather conditions and make their own decisions about what to do next.

Figure 1