Remediation cost on many Munitions and Explosives of Concern (MEC) sites directly relates to the amount of ferrous clutter on the surface and upper band (0-8”) of terrain. It is common to remove tons of ferrous scrap for each positive MEC find. The surface clutter is laborious to collect while the clutter remaining in the upper soil band results in false positives (digs). Each dig is costly and slows the MEC remediation process. While occasional MEC does lay on the surface and in the upper band of soil, at many sites the MEC is 12”-36” deep. It is estimated that MEC remediation cost per acre at many sites could be reduced by 30% or more if surface and near surface ferrous clutter could be efficiently removed before the final mapping and digging process commences. MEC remediation project speed would improve by dramatically shortening the most labor and time-intensive process—that of discriminating and digging.

This project was focused on developing and demonstrating the Semi-Automated Ferrous Material Scouring System (SAFMSS) that could be utilized for the remediation of large-scale range clearing projects. The major goals associated with the development of this system were to:

  • Increase the speed of range clearing operations
  • Improve personnel safety by removing workers from dangerous areas
  • Decrease cost to the government by decreasing personnel costs and days to completion

Technology Description

Caterpillar retrofitted the control system of a heavy duty excavator to enable either manual control or control via a remote computer. Sensors, wiring harnesses, electronic control units, and mounting hardware were added to complete the control system. Four high-resolution color cameras housed in environmentally hardened enclosures were mounted on the top of the operator cab to provide the video feedback for teleoperation.

A remote control station was constructed based on ruggedized personal computing technology and flat screen displays. High resolution images from the excavator mounted cameras are provided on three separate displays arranged to give a forward looking view. These displays provide the color video for terrain sensing.  A laser scanner and vision system are integrated to provide 3D data for the Situation Awareness with Colorized Ranging (SACR) software. As the remote operator swings the excavator boom 180 degree, the sensor provides the necessary range and video data to enable generation of a synthetic operating environment, terrain undulation, and obstacle detection.  Production joysticks and electronic control units were used for operation of the excavator. All of these components are mounted in a towed trailer for easy transport and protection from the elements. 

A multipurpose tool for the excavator called the Scarifying Magnet (SCARMAG) was developed to improve the way surface and near surface ferrous MEC is removed from ranges.  The SCARMAG end effector is an integration of a standard excavator bucket, thumb gripper, electromagnet and scarifying teeth. The SCARMAG consists of a bucket, a 16KW electromagnet having an effective diameter of 48”, an actuated thumb utilizing the excavator’s auxiliary hydraulic circuit for actuation and “teeth” added for scarifying of the soil. The SCARMAG can be used to 1) remove large brush or obstacles, 2) to push aside smaller brush and obstacles, 3) to flip the electromagnet into position and sweep the area to collect surface and near surface ferrous debris and 4) to use the teeth to loosen the soil in areas identified as potentially having significant subsurface clutter and then using the magnet to clear that particular area.

Demonstration Results

Due to the fact that the system magnetized the soil, traditional methods (metal detection instrumentation) would not work for validation of removal efficiency. In addition, the UXO techs supporting our effort were not permitted to disturb soil to find any items missed. Because of this quantitative data is not available for analysis.

Speed and coverage is validated by data logs recorded and stored at the remote control station. The primary method of validating the system’s ferrous debris removal performance was through review of video footage and walking the area after SAFMSS processing making visual observations. The five team members, including the 2 UXO technicians would visually scour the area for obvious missed items and compared that total with the total removed by SAFMSS. It is estimated that SAFMSS can remove between 80% and 90% of all ferrous surface debris and between 40% and 50% of all ferrous subsurface debris that are not filled with non-ferrous material (e.g. a steel container filled with dirt). The subsurface debris prove to be the most difficult to assess during the final test. The above estimate is based on visual observations of items that were visible but partially buried during the final test, and on system validation using seeded items during on-site testing.

Implementation Issues

During the project, a few issues were discovered that would need to be corrected to enhance the performance of the system and enable operational deployment.

  1. Magnetization of the soil makes traditional instrumentation based on magnetic field detection unusable. Alternative instrumentation such as GPR or other non-magnetic based instrumentation must be used in conjunction with a SAFMSS system to complete a remediation program. It was not determined if the magnetization of the soil decays over time or not.
  2. Training of operators and maintainers is needed to keep costs in line with the savings projections. It only takes 1 person to operate the system, but typically there would be.
  3. Since SAFMSS is a hydraulic powered system, spills during remote operation could become a major issue without some sort of spill sensing. To our knowledge there is no automated system provided by OEM excavator manufacturers to detect fluid leakage remotely.
  4. The SAFMSS had scarifying teeth that were 18” long. Scarifying teeth 24” long would be significantly more effective at breaking up soil to the desired 8 to 12 inches depth.
  5. An independent pan-tilt-zoom high-resolution camera for the system operator to remotely identify areas/objects of concern/interest would be needed in the final implementation. Every UXO technician and observer of the system in observation made this comment.

All technologies and sensors are available off-the-shelf but this was the first custom implementation and integration of all these technologies for this application. Hardening and commercialization of the independent components would be the next step.