0
Select Articles

Mechatronics for Humanitarian Explosive Ordnance Disposal in Cambodia PUBLIC ACCESS

 

Mechanical Engineering 140(09), S4-S10 (Sep 01, 2018) (7 pages) Paper No: ME-2018-SEP6; doi: 10.1115/1.2018-SEP6

Over the past six years, researchers at Villanova University (VU) and the Golden West Humanitarian Foundation (GWHF) have developed an integrated research and educational program focused on the use of mechatronics and robotics in humanitarian explosive ordnance disposal (EOD) and mine action. In the following article, I will talk about this program, discuss two ongoing projects - a low-cost EOD robot and an automated ordnance identification system - and talk about how we have successfully integrated students in the work. There are many opportunities for the DSCD community to get involved in this area and hopefully this article will pique your interest.

Over the past six years, researchers at Villanova University (VU) and the Golden West Humanitarian Foundation (GWHF) have developed an integrated research and educationalprogram focused on the use of mechatronics and robotics inhumanitarian explosive ordnance disposal (EOD) and mineaction. In the following article, I will talk about this program,discuss two ongoing projects – a low-cost EOD robot and an automatedordnance identification system – and talk about how we have successfullyintegrated students in the work. There are many opportunities for theDSCD community to get involved in this area and hopefully this article willpique your interest.

During the last week of spring classes in 2012, I was standing at the Villanova Mechanical Engineering copier, making copies of the finalexam for my Mechatronics class. Jordan Ermilio, director of the Villanova

Engineering Service Learning (VESL) program came out of his office and said, “Garrett, do you want to go to Cambodia?” I enthusiasticallysaid, “Yes!” This caught him by surprise and asked if I should checkwith my family first. This was a good point, so I did. They were insupport of the trip, and 3 weeks later I was on my way to Phnom Penhwith Jordan and a group of 3 students to lay the groundwork for a verynon-mechatronic construction project. As luck would have it, on theflight from Taipei to Phnom Penh, one of the students traveling withus sat next to Len Austin, GWHF’s EOD Special Operations Chief. He invited us to visit GWHF’s testing facility, we accepted, and over dinner with Len and Allen Tan, Director of Technology at GWHF, wedecided to collaborate on a low-cost EOD robot (discussed in moredetail below). Six years later, I have spent more than 6 months inCambodia working on a broad range of projects focused on developingmechatronic and robotic solutions to issues in humanitarian mineaction and EOD. For some context about why this work is done inCambodia, see Sidebar 1: “Cambodia’s Recent History.”

S1: Cambodia’s Recent History

Starting in 1965 and continuing for three decades, Cambodia was embroiled in armedconflict. US bombings during the Vietnam Warmade Cambodia one of the most bombed countries in history [1]. Many bombs did not explode,leaving un-exploded, but still active bombsthroughout the country (US bombing sites areshown in yellow/purple in Figure S1 along theCambodia-Vietnam border.) Following this period,the Khmer Rouge came to power, resulting in theCambodian genocide [2]. The Khmer Rouge wereeventually pushed to the Cambodia-Thailand border, where landmines were deployed by Vietnam,Cambodia, and Thailand to prevent the KhmerRouge from entering their respective countries [2](active landmine sites are shown in red in Figure S2 along the Cambodia-Thailand border).

FIGURE S1 Map showing unexploded ordnance (UXO) in Cambodia from theCambodian Mine Action Centre’s “PeaceMuseum for Mine Action” in Siem Reap,Cambodia. Legend: Yellow/Purple - US Bombing Sites, Red - Minefields.

Grahic Jump LocationFIGURE S1 Map showing unexploded ordnance (UXO) in Cambodia from theCambodian Mine Action Centre’s  “PeaceMuseum for Mine Action” in Siem Reap,Cambodia. Legend: Yellow/Purple - US Bombing Sites, Red - Minefields.

In Cambodia, more than 1970 km2 of land is contaminated by explosive remnants of war(ERW) [2]. During the period 1979 to 2017, therehave been 64,688 recorded casualties associatedwith mine/ERW [2-4]. Economic growth is stifledbecause expansion of farms and other livelihoodactivities have been impeded [2]. Thus, thereis a critical need to remove these threats. Morebroadly speaking, it is estimated that 100 million landmines are spread throughout 84 countries andthat landmines kill or injure 20,000 people eachyear, worldwide [5]. This problem is compoundedby the fact that many of these ERW are in low-income countries that cannot afford technologiesthat make removal easier and faster [5, 6].

S2: The Golden West Humanitarian Foundation

One noted difficulty with working in humanitarian robotics is access toexpertise [7]. Our work does not suffer from lackof access to expertise because of our strongand continuing collaboration with the GoldenWest Humanitarian Foundation (GWHF), a US-based NGO recognized by the internationalhumanitarian mine-action community as one ofthe premier non-governmental organizations forthe research, development, and implementationof solutions to address this sector’s most difficulttechnical challenges. GWHF brings EOD expertise,highlighted by an interdisciplinary staff of expertsconsidered to be one of the best in the fields of EOD, detection technologies, and engineeringresearch and development. In addition, GWHF has a long history of product development exemplified by their extremely successful 3D printed EOD training aids, shown in Figure S2.

Figure S2 GWHF lead engineer, John Wright, holding one of GWHF’s 3D printed EOD training aids.

Grahic Jump LocationFigure S2 GWHF lead engineer, John Wright, holding one of GWHF’s 3D printed EOD training aids.

Before we begin discussing mechatronics in humanitarian EOD and mine action, it isimportant to make the distinction between militaryand humanitarian missions. The goal of military EODis to render a threat safe and/or create safe routesfor soldiers. This requires, for example, only partialclearance of mine fields. In contrast, humanitarian operations focus on completely removing threats inorder for the land to be used safely – in the minefield example above, 100% clearance is desired. One could argue that the latter is the more difficult task, which underscores the need for advanced engineering fields, like mechatronics, to be applied in humanitarian EOD and mine action.

Partially the answer to the question above is about perception. When one thinks of “humanitarianengineering,” thoughts of “low-tech” or “old-tech” mightcome to mind. These thoughts are in direct conflict with mechatronics, a high-tech, cutting-edge field. This dissidence between humanitarian engineering and mechatronics might leave any engineer worryingthat robustness and/or safety have been traded foraffordability. This trade-off between robustness/safety and affordability, however, is diminishing asthe cost and availability of components necessaryfor mechatronic devices become cheaper and morewidely available. Consider that relatively high-power computer boards (e.g., Raspberry Pi) and relatively easy-to-program microcontroller boards (e.g., Arduino)are both available for under $10. These componentsare even in-stock at stores in low-income countries provided you know where to look, making developmentsin mechatronics relevant and important in low-incomecountries.

More generally speaking, mechatronic systems are used quite broadly in EOD, especially in higher-income countries. Systems that can detect, identify, handle, move, disarm, destroy, manipulate, or, in general, observe and/or interact with ordnance while keepinghuman operators safe are of the utmost interest to the EOD community. Currently, there are two key areas within mechatronics that have made significant impacts in this community – robotics and sensing.

Robots are widely used in EOD, allowing operators to interact with ordnance without direct contact [8, 9].There has also been a great deal of work on landmine—|robots [10, 11]. Although these are not currently deployed in Cambodia there are a number of new initiatives and companies aimed at developing robots for mine action in Southeast Asia.

FIGURE 1 The VU-GWHF EOD robot performing a simulated EOD mission. Photo taken at the GWHF test facility in Kampong Chhnang, Cambodia.

Grahic Jump LocationFIGURE 1 The VU-GWHF EOD robot performing a simulated EOD mission. Photo taken at the GWHF test facility in Kampong Chhnang, Cambodia.

Ordnance sensing (e.g., detection) is also an active mechatronics-related research field which has had animpact in humanitarian demining. Research in this area is focused on signal processing and statistical methods,as well as sensor design and development [12]. Novelinstruments are constantly being developed like the new Scorpion unexploded ordnance (UXO) detector developed by the US DoD through the Humanitarian Demining Research & Development (HDR&D) program.

Along with our project partner GWHF (see Sidebar 2: “The Golden West Humanitarian Foundation” for more information), we have been working on a broad range of problems in humanitarian EOD and mine action. Everything from the mechatronics-related projects I will discuss in more detail below, to life-cycle analysis studies on GWHF’sExplosive Harvesting program to field-deployable fume hoods to lessen the environmental impact of ordnance disposal. In this section, I am going todiscuss, in brief, two ongoing projects.

After that first chance meeting, the VU-GWHF team began work on a low-cost EOD robot. Robots are widely used in EOD, allowing operators to interact with Ordnance, e.g., an improvised explosive device (IED),without direct human contact. Typical EOD robotshave a robotic manipulator with varying degrees of freedom (multi-link arms with versatile grippers), anumber of cameras, are able to traverse diverse terrain, and are remotely-operated. Many different robots havebeen developed, for example military-grade robotslike the Talon [9] or the PacBot [8]. Even though therehas been a large amount of work in EOD robotics,the current commercially-available robots are too expensive for use in low-income countries. Note that even when initial purchase prices can be lowered, the maintenance costs can be significant, presenting an additional barrier.

The VU-GWHF low-cost EOD robot [13] can be seen in Figure 1. Although I could discuss the many technical challenges we encountered while developing this robot, perhaps the most interesting difficulties we encountered were centered on designing for useand/or manufacture in low-income countries. When we first approached this problem, we put cost as our most important design constraint, which led to a robot constructed out of PVC and parts that could be entirely sourced in Cambodia. While this makes sense becausecost and supply chain are major concerns, at our first design review, GWHF made it clear that PVC wouldn’tcut it. Reliability/robustness and functionality were,in some cases, more important than cost and we could justify spending more on and importing critical parts and materials (although cost could never be forgotten).

Once the interplay between these three main design considerations (cost, reliability, and functionality) was established, we set out to produce an EOD robot that had 90 percent of the functionality of military-graderobots at 10 percent the cost. This led to tradeoffs in a number of the robot’s subsystems, which leaves it different from typical military EOD robots in a numberof ways as discussed below.

  1. 1. Robotic Manipulator: The robot has a relatively-simple 2DOF arm, which stands in stark contrast to commercial robot arms, which can be highly articulated [8]. This choice was made to reduce cost and increase robustness and field repairability. In consultation with GWHF EOD specialists, it was decided that this arm would still be useful in many EOD missions. Note that the linear actuators that drive the arm are high-end actuators that can withstand the harsh environments in which the robot will be deployed – an example of when high-cost components can be justified.

  2. 2. Drive System: The robot’s drive system uses wheels, not tracks. Wheels were chosen to reduce cost and maintenance but limit the robot’s ability to handle difficult terrain. This design decision was perhaps the most difficult to make because treads would vastly expand the robot’s capabilities. In this case, cost and reliability outweighed functionality.

  3. 3. Electronics and Communications: The robot’s control electronics are almost entirely chosen from maker-movement technologies, i.e., Arduino, Raspberry Pi and lower-cost components. Note that these low-cost computationalcomponents do not allow the flexibility in programming of higher-cost, higher-power targets, resulting in a trade-off between functionality and cost. Wireless communication was identified as a critical subsystem, so higher-cost components (resulting in higher functionality and reliability) were justified.

Current Status: Current efforts on this project are focused on developing a production prototype. We are hoping to have these robots available for field work inthe next year.

FIGURE 2 Example images of mortars showing, from left to right. Top row: Yugoslavian M47 HE, Chinese Type 22, Chinese Type 30, USSR S832SM Illumination. Bottom row: Vietnamese Propagande Model UNK, USA M374 series HE, USSR F-843

Grahic Jump LocationFIGURE 2 Example images of mortars showing, from left to right. Top row: Yugoslavian M47 HE, Chinese Type 22, Chinese Type 30, USSR S832SM Illumination. Bottom row: Vietnamese Propagande Model UNK, USA M374 series HE, USSR F-843

FIGURE 3 The 2015 EOD robot team during field work in Cambodia at GWHF’s Kampong Chhnang testing facility. From left to I right. Back row: Chenda Nget (GWHF Engineer), John Wright (GWHF Lead Engineer), John Lim, Denzen Boldsikhan, Anthony Marone, Ian Stankosh, Sombo Heng (GWHF Programmer), Christopher Hsu, Samantha Tropeano, Dim Samach (GWHF Engineer). Front row: Garrett Clayton, the EOD robot, Allen Tan (Director of GWHF’s Design Lab).

Grahic Jump LocationFIGURE 3 The 2015 EOD robot team during field work in Cambodia at GWHF’s Kampong Chhnang testing facility. From left to I right. Back row: Chenda Nget (GWHF Engineer), John Wright (GWHF Lead Engineer), John Lim, Denzen Boldsikhan, Anthony Marone, Ian Stankosh, Sombo Heng (GWHF Programmer), Christopher Hsu, Samantha Tropeano, Dim Samach (GWHF Engineer). Front row: Garrett Clayton, the EOD robot, Allen Tan (Director of GWHF’s Design Lab).

A more recent project, which is co-led by Dr. C. “Nat” Nataraj at VU, focuses on the development of an automated ordnance classification system to aid EOD technicians [14]. In typical EOD missions, classification of the ordnance that is being dealt with is of critical importanceas it dictates how the EOD technician will approach the situation. High-confidence classification requires anexpert with a great deal of training and experience. This expertise is not always available in low-income countries, so this project seeks to aid EOD technicians in low-income countries in the ordnance classification process through the use of machine learning.

The concept for this classification support system is that an EOD technician will take a picture of a piece of ordnance and the system will return potential ordnance matches using a ConvNet classifier [14]. While these techniques are well established in machine learning, the availability of ordnance image datasets is limited or non-existent. Thus, the firststep in this project was to acquire such a dataset. This dataset is composed of images taken from ordnance libraries belonging to GWHF, the Cambodia Mine Action Centre (CMAC), and other libraries in Europe.

Our growing image database currently has more than 250 different types of ordnance. Example images of mortars can be seen in Figure 2.

Preliminary results have been promising. As reported in Ref. [14], classification of mortars from the set in Figure 2 was found to be 97% accurate. Current workis focused on expanding the image database, testing the methods on additional types of ordnance, and enabling the classification of partially-occluded ordnance andordnance in field conditions.

About the Author

 

Grahic Jump Location 

Garrett M. Clayton is an Associate Professor in the Department of Mechanical Engineering and the director of the Center for Nonlinear Dynamics and Control at Villanova University.

His current research interests are broadly spread throughout the area of mechatronics with specific applications in nanopositioning, remote monitoring, humanitarian technologies, and robotics. Dr. Clayton received his Bachelor of Science in Mechanical Engineering from Seattle University in 2001 and his Masters of Science and PhD in Mechanical Engineering from the University of Washington in 2003 and 2008, respectively. His research has been funded by the National Science Foundation (NSF), the US Environmental Protection Agency (USEPA), the Pennsylvania Department of Transportation (PennDOT), and the Golden West Humanitarian Foundation (GWHF) and a number of small companies. He recently received a Fulbright Research/Teaching grant to live and work in Cambodia in the summer of 2018.

During the course of this project, more than 60 students have taken part in these collaborative projects through senior design projects, summer internships and fellowships, undergraduate research, Master’s theses, PhD dissertations, and volunteer work. Of these 60 students, more than 20 have traveled to Cambodia for field work (some for trips as long as twomonths). A few of these students are shown in Figure 3 with some of our collaborators at GWHF. Anecdotally, these international project experiences have affected the students in a meaningful way: more than 10 have entered the armed forces (3 in the Navy nuclear program’s highest engineering position and at leastone in EOD) and more than 15 have gone on to pursue advanced degrees in engineering. We will continue to involve a broad range of students through a recent NSF International Research Experience for Students (IRES) grant to expand summer fellowships.

The VU-GWHF research program discussed above is focused on the use of mechatronics and robotics to solve problems in humanitarian EOD and mine action. Two ongoing projects were presented that seekto address needs specific to low-income countries like Cambodia. As stated previously, there are a number of intriguing, difficult, rewarding problems to which the DSCD community could contribute solutions. If you are interested in these problems, I encourage you to reach out to me for more information. Collaborators are welcome!

This work could not have been possible without the many students who have taken part. I thank them all, but would like to specifically thank Michael Benson and Anderson Lebbad who have spent nearly as much time in Cambodiaas I have. I would also like to thank my colleagues, Dr. C. “Nat” Nataraj (VU), Dr. Jordan Ermilio (VU), Allen Tan (GWHF), and John Wright (GWHF), all of whom I have collaborated with very closely on these projects. Finally, I would like to thank our sponsors for their generous support. The initial trip to Cambodia was funded by the Caramanico Foundation (they continue to support projects in Cambodia to this day). GWHF, through a U.S. State Department Grant, has supported anumber of senior design projects, summer internships, and graduate theses and dissertations. Current work is funded through a Fulbright Scholar grant and an NSF IRES Award (Award Number 1658696).

Owen, Taylor and Ben Kiernan. “Bombs over Cambodia.” THIRD WORLD RESURGENCE 201 (2007): p. 41.
The Royal Government of Cambodia. “National Mine Action Strategy 2018-2025.” 12 December 2017.
Sambath Chan, “Munitions Risk Education in Cambodia”. The Journal of ERW and mine action. Spring 2013, pp. 38-42.
Carla Wheeler, “GIS Technology Helps Rid Southeast Asia of Dangerous Land Mines and Unexploded Ordnance”. ArcWatch: Your e-Magazine for GIS News,Views, and Insights. July 2008.
Portugal, David, Lino Marques and Manuel Armada. “Deploying Field Robots for Humanitarian Demining: Challenges, Requirements and Research Trends”. Proc. of the 17th International Conference on Climbing and Walking Robots (CLAWAR 2014), Poznan, Poland. 2014.
Siegel, Rob. “Land mine detection.” Instrumentation & Measurement Magazine, IEEE 5.4 (2002): pp. 22-28.
Trevelyan, James. “Robots and Landmines.” Industrial Robot: An International Journal, 24(2) 1997, pp. 114-125. [CrossRef]
Yamauchi, Brian M. “PackBot: a versatile platform for military robotics.” Defense and Security. International Society for Optics and Photonics, 2004.
Wells, Peter and Dan Deguire. “TALON: A universalunmanned ground vehicle platform, enabling the mission to be the focus.” Defense and Security. International Society for Optics and Photonics, 2005.
Yvan Baudoin and E Colon. “Humanitarian demining and robotics” In Proceedings of the 1998 IEEE International Conference on Control Applications,” volume 1, IEEE, 1998, pp. 433-435.
S Rajasekharan and Chandra Kambhampati. “Thecurrent opinion on the use of robots for landmine detection”. In Proceedings of the IEEE International Conference on Robotics and Automation, volume 3, IEEE, 2003, pp. 4252-4257.
Gooneratne CP, Mukhopahyay SC and Sen Gupta G “A review of sensing technologies for landmine detection: Unmanned vehicle based approach”. In Proceedings of the 2nd International Conference on Autonomous Robots and Agents, Palmerston North, New Zealand, 2004, pp. 401-407.
Fracchia M., Benson M., Kennedy C., Convery J., Poultney A., Anderson J.W., Tan A., Ermilio J., Clayton G.M. “Low-Cost Explosive Ordnance Dispose Robot for Deployment in Southeast Asia”. 2015 IEEE International Humanitarian Technology Conference. May 2015.
Lebbad, Anderson, Garrett Clayton and Nataraj C. “Classification of UXO Using Convolutional Networks Trained on a Limited Dataset.” 2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE, 2017.
Copyright © 2018 by ASME
View article in PDF format.

References

Owen, Taylor and Ben Kiernan. “Bombs over Cambodia.” THIRD WORLD RESURGENCE 201 (2007): p. 41.
The Royal Government of Cambodia. “National Mine Action Strategy 2018-2025.” 12 December 2017.
Sambath Chan, “Munitions Risk Education in Cambodia”. The Journal of ERW and mine action. Spring 2013, pp. 38-42.
Carla Wheeler, “GIS Technology Helps Rid Southeast Asia of Dangerous Land Mines and Unexploded Ordnance”. ArcWatch: Your e-Magazine for GIS News,Views, and Insights. July 2008.
Portugal, David, Lino Marques and Manuel Armada. “Deploying Field Robots for Humanitarian Demining: Challenges, Requirements and Research Trends”. Proc. of the 17th International Conference on Climbing and Walking Robots (CLAWAR 2014), Poznan, Poland. 2014.
Siegel, Rob. “Land mine detection.” Instrumentation & Measurement Magazine, IEEE 5.4 (2002): pp. 22-28.
Trevelyan, James. “Robots and Landmines.” Industrial Robot: An International Journal, 24(2) 1997, pp. 114-125. [CrossRef]
Yamauchi, Brian M. “PackBot: a versatile platform for military robotics.” Defense and Security. International Society for Optics and Photonics, 2004.
Wells, Peter and Dan Deguire. “TALON: A universalunmanned ground vehicle platform, enabling the mission to be the focus.” Defense and Security. International Society for Optics and Photonics, 2005.
Yvan Baudoin and E Colon. “Humanitarian demining and robotics” In Proceedings of the 1998 IEEE International Conference on Control Applications,” volume 1, IEEE, 1998, pp. 433-435.
S Rajasekharan and Chandra Kambhampati. “Thecurrent opinion on the use of robots for landmine detection”. In Proceedings of the IEEE International Conference on Robotics and Automation, volume 3, IEEE, 2003, pp. 4252-4257.
Gooneratne CP, Mukhopahyay SC and Sen Gupta G “A review of sensing technologies for landmine detection: Unmanned vehicle based approach”. In Proceedings of the 2nd International Conference on Autonomous Robots and Agents, Palmerston North, New Zealand, 2004, pp. 401-407.
Fracchia M., Benson M., Kennedy C., Convery J., Poultney A., Anderson J.W., Tan A., Ermilio J., Clayton G.M. “Low-Cost Explosive Ordnance Dispose Robot for Deployment in Southeast Asia”. 2015 IEEE International Humanitarian Technology Conference. May 2015.
Lebbad, Anderson, Garrett Clayton and Nataraj C. “Classification of UXO Using Convolutional Networks Trained on a Limited Dataset.” 2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE, 2017.

Figures

Tables

Errata

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In