Landmines Detection by Using Mobile Robots

List of Contents

Chapter 1: Introduction
1.1 Landmines over the world
1.2 Challenges
1.3 Motivations
1.4 Contributions
1.5 Thesis Outline

Chapter 2: Landmines detection strategies
2.1 Landmines Detection Overview
2.1.1 Classification of Mines
2.2 Mine Detection Technologies
2.3 Metal Detector
2.4 Electromagnetic Methods
2.4.1 Ground Penetrating Radar (GPR)
2.4.2 Nuclear Quadruple Resonance (NQR)
2.4.3 Microwaves
2.4.4 Electrical Impedance Tomography(EIT)
2.4.5 Infrared Method
2.4.6 X-Ray Backscatter Method
2.4.7 Sound and Ultrasound
2.4.8 Neutron Method
2.5 Acoustic/Seismic
2.6 Biological Method
2.6.1 Dogs and Rats
2.6.2 Bees
2.6.3 Bacteria
2.6.4 Antibodies method
2.6.5 Chemical Methods
2.7 Mechanical Methods
2.7.1 Probes and Prodders
2.7.2 Mine Clearing Machines
2.8 State of discussed solutions
2.9 Robot detecting strategies
2.9.1 Disadvantages of Robots
2.10 Summary

Chapter 3: Demining robot techniques
3.1 Motion planning
3.1.1 Motion planning Problem
3.1.2 Locomotion
3.1.3 Motion system
3.1.4 Motion Planning T echniques
3.1.4.1 Sampling-Based Planning
3.1.4.2 Probabilistic Roadmap Method
3.14.3 Graph Search
3.1.4.4 Uninformed Search
3.1.4.5 Heuristic Search
3.1.4.6 A* Algorithm
3.1.4.7 Complete coverage Technique
3.1.5 Area Mapping
3.2 Sensor Fusion
3.2.1 Sensor fusion Applications
3.2.2 The multiple sensors integration
3.2.3 Multiple Sensors Fusion Advantages
3.2.4 Sensor fusion problems
3.2.5 Multisensor Fusion
3.2.5.1 The Integration F unctions
3.2.5.2 The Rule-Based and Network System
3.2.6 Multisensor Fusion Levels
3.2.6.1 Signal-Level Fusion
3.2.6.2Pixel Level Fusion
3.2.6.3 Feature-Level Fusion
3.2.6.4Decision Level Fusion
3.2.7 Sensor fusion system in landmines detection

Chapter 4: Related Work
4.1 Landmines Detection by Robots
4.2 Summary

Chapter 5: The proposed framework
5.1 The Proposed Low-Cost Robot System
5.2 Motion Planning
5.3 Low cost robot structure
5.4 Multi-sensor Fusion (Decision Level Fusion)
5.5 The vision system
5.6 Destroying mines
5.7 Conclusion

Chapter 6: The Experimental Results
6.1 System Components
6.1.1 Chemical sensor
6.1.2 Metal Detector
6.1.3 Ultrasound Sensor
6.1.4 Camera Sensor
6.2 Experimental Results
6.3 Discussion
6.4 Conclusion

Chapter 7: Conclusions
7.1 Conclusions
7.2 Future Work
Published papers

References

Acknowledgement

First and foremost, gratitude is due to almighty Allah, whose mercy has guided me the whole way through.

I would like to express my deepest admiration to Professor Hazem ElBakry, the head of information system department, Mansoura University. I have been trying to put my most profound appreciation into words but found no words to match my gratitude to his encouragement, careful follow up of this work, honest supervision, and close guidance.

Special thanks to Dr. Mohammed Elmogy, department of Information Technology, Faculty of Computers and Information, Mansoura University, who has always been there to help me overcome the obstacles I encountered during my research. Thanks for his responsiveness from the very beginning until the end, and the huge amount of time he always had for me.

My profound gratitude to my family for their continuous help and encouragement. I am most grateful for all of them and I am richly blessed to have been one of them.

Ahmed Ismail 2016

Abstract

Landmines are dangerous problem that threats people lives and causes economic problems. Landmines are harmful because of their unknown locations and difficulty to detect them. There are more than active 100 million landmines in the world. Detecting and clearing mines demands specific expertise. There are many techniques were used to detect landmines and clear them. The robotic method was an effective technique that solved the risks on human operators.

This thesis studies strategies for humanitarian demining using robotic units. We present a low-cost system for landmines detection. The proposed system uses fusion of low cost multi sensors instead of using very expensive one. The proposed robot used sensor fusion technique to increase the probability of mine detection. We have developed decision level fusion to decrease false alarm of mines detection. We used complete coverage path planning to find all possible mines in the environment.We proposed using multiple robots with the same structure to use complete coverage path in parallel way to save the time.We proposed effective obstacle avoidance algorithm to help the robot moves in autonomous motion.The proposed robot is light in order no to trigger mines and be destroyed. We proposed effecive method to destroy mines where they are using arm on the robot to help defusing method.

The purpose of the thesis is to give an efficient solution for the landmines problem. By using robots that are capable of exploring and destroying buried landmines. We also aimed to make the proposed robot with simle components to provide the soldiers and local landmines evivronments citizens with effective solution that they can use to save their lifes.

The information systems concerned with the systems which collects the available information from different sources to support the final decision.So, the proposed system worked on integration of multi sensors data then takes a decision wether there is amine or not.The propsed system is evaluted using simulated environment as well as areal landmines field with many scenarios. The results proved that be able to identify landmines types and clear them under different conditions with a succes over than 90%.The resullts can be better than achieved results by using more effecive sensors whichs are available in local markets.

List of Abbreviations

Abbildung in dieser Leseprobe nicht enthalten

List of Figures

Abbildung in dieser Leseprobe nicht enthalten

List of Tables

Abbildung in dieser Leseprobe nicht enthalten

Chapter 1 Introduction

Landmines are weapons that are used to prevent the attack or to deny access to military and civilians to a particular area.Landmines are placed on the ground or underground to destroy an area when they are triggered by humans, animals, or machines. Landmines are active for a long time until they are triggered [1]. It means that mines kill tens of people every day in locations where they exist [2]. Antipersonnel mines are an international threat to civilians. Because of mines, there are 15,000- 20,000 suffers all over the world per year. The activities such as farming, manufacturing and trading are effected at mine fields.

The landmines are two types:antipersonnel mines and antitank mines. Antipersonnel mines are located under the earth and close to the surface while antitank mines are usually found on the surface of the earth. Landmines can e located everywhere either undergroung or on the ground.They could be planted by soldiers,thrown by vehicles or by planes in an ordered or disodrdered manner.Despite the funding from many nations and organizations, according to the current clearing rate, all existing mines need 450-500 years to be cleared.

1.1 Landmines over the world

There are an estimated 110 million active landmines buried in over 64 countries around the world.Landmines are in different environment conditions.They are in desert areas (Egypt, Kuwait, Somalia), in mountains(El Salvador, Afaghanistan), in jungles(Vietnam,Cambodi), and in urban areas(Former Yougoslavia,Beirut) .Table 1.1 and figure 1.1 shows landmines distribution in the world [2].

Abbildung in dieser Leseprobe nicht enthalten

Fig 1.1: The existence of the mines [3].

Table 1.1 mines over the world [2].

Abbildung in dieser Leseprobe nicht enthalten

Egypt was declared by landmine monitor in late 2005 as “one of the most mine-infested countries in the world.” The country has been reported to have approximately 20 million landmines and unexploded ordnances (UXOs) on its territory.The presence of landmines hinders significant agricultural and construction projects, making it difficult to establish industrial cities in many areas. The minefields in the region, known as the "Devil's Gardens," were planted by German commander Erwin Rommel as a line of defense against British troops who occupied Egypt during World War II. The past seven years alone, the number of victims of landmines was 736 people, 43 of whom lost their lives.

1.2 Challenges

There are many challenges that made the process of demining as a very difficult process because of one or more of the following causes :

1- There are many changes in weather conditions that led to the change and disappearance of the landmines positons.
2- The difficulty to defusing mines without an vision underground.
3- The location of landmines should be defined with an accuracy of the 5cm radius.
4- The antipersonnel landmines detection costed between $3-$25 per mine and a mine clearance costed between $300-$1000 [1].
5- The ability of robot locomotion in mines filed.
6- The high cost of demining robots
7- The landmines fied maps are not accurate ,so mines defusing is not accurate.

1.3 Motivations

We developed our low cost demining robot for the following purposes:

1- Every landmines detection technique has error rates,therefore using combined techniques by using sensor fusion technique will decrease error rates.
2- Despite the world community efforts to end the landmines problem in sort time,the demining is not changing fast.This is due to the limited performance of available techniques.
3- In the last years,many demining researches used ground penetrating radar as the best technique to detect mines.However, ground penetrating radar is very expensive ,very heavy ,and power hungry.Therefore ,using fusing of cheap sensors to detect mines can solve ground penetrating radar problems.

1.4 Contributions

The research’s aim is to develop a mobile robot system that is identifying and destroying landmines. One of the most critical issues in landmines detection is defining where are mines exactly exist. There are many techniques to solve his issue with false alarms. The purpose of this research is to design a multi-sensor robot that collects information from the environment for detecting and defusing landmines in landmines field. The research activities are outlined as follows:

1- Desinig a low-cost autonomous mobile robot that uses effective motion planning in mines detection.
2- Using multisensor fusion to combine more than one technique to decrease landmines detection error rates.
3- Developing effective object avoidance algorithm to pass through the field.

Using fusion of demining sensors to enhance its performance, as every sensor has some false alarm errors in detection.

4- Using effective technique to defuse the mines where they are instead of forming maps with high error rates.

1.5 Thesis Outline

The thesis is organized into seven chapters as follows:

Chapter 2 introduces the strategies of landmines detections as applied in the field of landmines detection. It describes the new methods in mines detection showing their power compared to classical methods.

Chapter 3 discusses the different robot motion planning techniques and how can robot apply defined motion planning in landmines detection.It introduces also,the data fusion method and how the fusion information can be be used to decrease the falas alarms in mines detection.

Chapter 4 discusses the previous related work in landmines detection by robots to represent the efficient solutions for demining.

Chapter 5 presents the proposed framework. At the beginning of the chapter, it draws a general view of the development environment, which contains the robot design. Then it illustrates implementation and using coverage based algorithm as path planning to search for mines. Then the chapter describes how the robot's parts interact with each other to collect data from different techniques together to decrease false alarm and increase the probability of mines detection. Then it shows how the low-cost autonomous robot can use defused data to make the mines detection ae more effective.

Chapter 6 presents the results and conclusion about the results.It shows how the low cost autonomous robot can detect mines with high accuracy as same as the advanced very expensive robots.

Chapter 7 presents the future work to enhance the robot for better demining.It suggests to use new techniques like nanotechnology and orgone energy to detect mines

Chapter 2 Landmines detection Strategies

There are many landmines detection techniques had been developed and tested in different environment conditions. Many sensor types like ultrasound sensor, ground penetrating radar, infrared sensor, camera sensor, and chemical sensor, are used to detect mines at different depth. Electrical Impedance Tomography(EIT) technology uses conductivity under the landmines field to detect landmines. The ground penetrating radars use high frequency of electromagnetic waves to obtain images about the underground components to detect mines. The explosive vapor in landmines is used as evidence of the existence of the mines by using chemical sensors. The development of robot to detect mines could help in risky landmines detection by using landmine sensors. Sensors collect precise data from the landmines field to discover anomalies by signals and image processing methods to detect mines.

This chapter presents the old and new techniques in landmine detection field. We here study different strategies for landmines detection to know all available techniques. The chapter introduce every method that can be used to detect landmines in order to show the property that helps in the detection method and show its limitations and strengths.

2.1. Landmines Detection Overview

There are two kinds of landmines: Anti-Personnel(APM) Mines and Anti­Tank (AT) mines. AP mines usually located under the earth and close to the surface while AT mines are typically located on the surface of the earth. Landmines detection and clearing is very risky and costly process. Searching for mines is a difficult task as there is no knowledge about the mines existence. There are about 2000 different types of mines for which catalogs exist [4]. As there is no standard for the final form of the mines. They can be like cans, rods, circles, and many other styles. Landmines field clearance needs two steps. The first one is to discover mine location. The second is to destroy or deactivate the mines. Locating mines process takes the longest time from the overall process. Since, every inch of the landmine field needs to be carefully inspected [5]. There are three strategies are used for landmines detection: a manual, a mechanical, and advanced robotics [6]. The manual technique uses trained men who sweep the field with metal detectors. Its speed is approximately 25 m2/h. As the landmines size is small and as many of mines are made ofnon-metallic materials, this method is expensive, slow, and dangerous.

The mechanical method relies on a machine that goes over the mines field and destroy them whether they are underground or on the ground. This method is the fastest method. However, the machines used in this method are so expensive to operate and used only when the terrain is suitable. Using mobile robot for mines detection ensures the safety of deminers and soldiers. The robot sweeps the mines field to discover the mines. The robot is sensitive and reliable. It has low-cost accurate sensors that will help in landmines detection.

2.1.1 Classification of Mines

Landmines are classified according to purpose into three kinds, antipersonnel mines (APM), antitank mines (ATM), and unexploded ordnance (UXO) as shown in Table 2.1

Table 2.1: the classification of the mines [3].

Abbildung in dieser Leseprobe nicht enthalten

The UXO are bombs that had been exploded but remain. ATM contains metallic material and is bigger than the size of APM. ATMs were used to destroy vehicles, so they are exploded by heavy weights on the ground. The ATM can be planted under the ground by specialists with different distances. The ATM can be used for large fields to stop the enemines for forward moving.

Abbildung in dieser Leseprobe nicht enthalten

Fig 2.1 Examples of ATM [6].

Table 2.2: Specifications of the ATM [3].

Abbildung in dieser Leseprobe nicht enthalten

APMs are harder to locate and clear. In addition, they harm the civilian people. APMs consist of nonmetallic material, and they have smaller size than the size of ATM. They need light pressure to be triggered. ATMs are three kinds: bounding fragmentation,blast, and directional fragmentation. Antipersonnel mines could be located in any place. The blast mines are smaller and lighter than other types.

The bounding fragmentation can easily destruct expanded field, but the blast landmines can destroy only the specific object through a small area. The bounding fragmentation can be exploded by direct pressure. When the trigger is activated, they can destroy an area from 30m till 200m in radius. Directional fragmentation landmines are usually on the soil, and they can destroy objects in one direction. As they are exploded by a manual operation like animals or people, they considered as an active weapons.

Table 2.3: Specifications of the APM [3].

Abbildung in dieser Leseprobe nicht enthalten

The PRB-M35 and the PMN are blast landmines, can be exploded or triggered by 8kg weight. The PRB-M35 is the smallest landmine in size with 6cm diameter like Coke can diameter. When they are underground or covered with vegetation, it will be very difficult to detect them. As they are very light so they can easily move with rain water, so their location is difficult to predict. The PMN is made of a nonmetallic mine. The PMN is a cheap. The Valmara-69 is a bounding fragmentation mine. When they are triggered, they can destroy 27m radius area. The MON-100 can spread their fragment in a particular direction within an area up to 100m and can cover a 9.5m arc.

2.2. Mine Detection Technologies

The effectiveness of any method is estimated by the deminers risks, false alarms and the success in mines detection rates. Various methods were used to detect landmines; Demining and clearing mines can be risky, slowly and costly process. Using an autonomous robot to detect mines will guarantee the residents and deminers safety [6]. The current technologies fall under six main areas [7]: metal detector technologies, electromagnetic techniques, acoustic/seismic techniques, biological techniques, mechanical techniques and latest techniques.

2.3. Metal Detector

The landmine can be detected by the emitted electromagnetic field caused by the metallic objects in the soil. The metal detector sensor is cheap and so easy to use But ,can't be used for plastic landmines detection [8]. Metal detector used two mehods,detecting metal objects in the ground and sends audio signal to the operator or forming an image from detected object . using image processing to form an image about the detected object will help in give more information rather than catching an audio signal. The metal objects can be detected from 1cm to a depth of 65cm.The sensor measures a magnetic field by emitting a signal around a metal rod that produces a magnetic field that penetrates the ground. The ground disrupts the magnetic field, which measured by the magnetometer [8].

Metal detectors identify all metallic objects but, they can not differentiate a mine or from other objects. The large quantities of metal objects and other metal debris in most battlefields lead to false alarms by percentage 100-1000 false alarms for each detected landmin. The False alarm interferes with really mine as they waste the deminers’ time and concentration [9].

2.4. Electromagnetic Methods

The Electromagnetic Technique works based on the change in the electromagnetic properties of the ground surface. For example, ground- penetrating radar (GPR)[10], electrical impedance tomography (EIT) [7], X-ray backscatter, and inffared/hyperspectral systems [11].

2.4.1 Ground Penetrating Radar (GPR)

GPR consist array of sensors; each sensor emits the electromagnetic wave. Then travel through the medium and they are reflected depending on properties of conducting medium [8]. Reflected electromagnetic waves are received by the receiver unit. to detect the presence of anomaly is signal strength, as it tends to deviate when it encounters object with different electrical properties with reference to conducting medium. GPR can provide both presence of an object and the location of the soil. GPR works in the same manner of ultrasonic sensors; the main difference is that GPR is from 100MHz to 100GHz operating frequency. Electromagnetic waves depend on properties of the soil like moisture, homogeneity, electrical properties. The knowledge of these parameters helps in heightens accurate detection. The GPR operating is easy. GPR Can find mines and form an image of the buried objects using dielectric constant variations.

GPR is one of the best methods for underground research. However, using GPR when the ground is irregular decreases the measurement efficiency. GPR measurements rely on the weather conditions, soil type and soil water content [12]. There are techniques to detect real mines by enhancing GPR performance. The main goal is capturing the difference between the landmine components and the ground components. The system should use a high frequency to very small plastic mines at shallow depths [13].

2.4.2 Nuclear Quadruple Resonance (NQR)

Nuclear Quadruple Resonance use signal-to-noise ratio (SNR) to obtain good results. NQR is not affected by soil conditions. The major weakness of using NQR because of its nuclear properties. As The mines have explosive components(TNT), the signal will be weaker than others do. Another problem is the interference between NQR and radiofrequency from the environment. Nuclear quadrupole resonance depends on observation of radio-frequency signals from the 14N nuclei present in explosives [7].

The frequency of explosive mines signals is between 0.5 and 6 MHz NQR provides a good identification and can estimate quantity or depth of mines. NQR is highly effective if the NQR sensor is not exposed to radio frequency interference.

2.4.3 Microwaves

The main idea is constructing image on the computer from the reflected signals from minefield by sending microwaves signals into the ground. The most advantage of using microwaves in landmines detection that it constructs computer images of the ground’s states, like stones, and mines. The weakness of this method that Ambiguous results produced by plastic mines [7].

2.4.4 Electrical Impedance Tomography(EIT)

The main idea using electrical signals to construct image the distribution of the medium underground. The method is suitable for all types of mines detection. It is suitable for wet environments, since, the conductivity of the moist substrate will be better. The Electrical Impedance Tomography device is inexpensive and simple. The disadvantage of this method that the need for physical contact with the landmine field that may detonate the mine. EIT cannot do well in dry soil. EIT cannot detect mines at high depth or moist environments. The advantages of this method that EIT is suitable for detecting all mines types as metallic and nonmetallic mines create conductivity anomalies; the EIT equipment is relatively simple and inexpensive.

2.4.5 Infrared Method

Infrared/hyperspectral detects different variations in electromagnetic radiation emitted by objects. There are two types of irradiation active and passive which depend on a range of electromagnetic wavelengths. The thermal method exploits differential in temperatures of mines areas in comparing to surrounding areas. The ground above landmines is warmer than surrounding areas during the day but lose heat at night [7]. The Laser illumination is used to discover the different in temperatures. The advantage of this method that there is no direct contact with landmines and can be done from a far safe distance. The infrared method is good at wide areas quickly scanning. The Infrared method is suitable for on ground landmines, but it is not suitable for underground landmines.

2.4.6 X-Ray Backscatter Method

This method depends on sending X-rays into the surface with lower atomic numbers, like scatter x-ray radiation because of the electron density. X-ray system is light and small. X-Ray backscatter is poor in penetrating the soil. If the source is on low strength to be safer for a person who port the system, it will take a long time to take an image. The technology is effective to detect mines without physical contact to ground. To obtain an image for antipersonnel mines, high spatial resolution about 1 cm is required.

2.4.7 Sound and Ultrasound

Ultrasound uses high-pitched sound waves to obtain an image of underground ultrasound detection. The image is produced by the sound waves emission that are reflected by using frequency higher than 20 kHz into the ground. The reflected waves returned from mines field have changed acoustical properties. The ultrasound signal penetrates the soil and detects buried mines signals. In the wet ground, this method operates well. Ultrasound has a problem with the interface with air and soil.

The acoustic and seismic system emits waves through speakers to obtain vibration over the ground. The reflected waves from the ground that may contain landmines. The change in frequency and amplitude help in mines detection. The advantage of this system that there is no contact with the surface. Another strength of this system that the error rate is very low; but, other metal objects like cans may mislead the system. The Ultrasound waves can penetrate medium even if there are a high degree of humidity and air is resistant whereas electromagnetic waves do not do well in a high degree of humidity.

2.4.8 Neutron Method

Neutron Method relies on the elements excitation in the soil. The main idea of this method that explosive objects emits gamma rays or neutrons. When using a low-strength radiation source to send a signal to the soil. If there are mines, they will emit gamma rays due to physical properties of neutron moderation. Ground- surface conditions and the sensor distance effect in imaging systems and increase the false alarm. The remotely-controlled vehicle has a head 300 to 400kg weight, but a manned vehicle would have to be double weight to protect the operator.

2.5. Acoustic/Seismic

This technique looks for landmines by vibrating mines with sound waves through emitting them into the ground. As the sound waves reflect different waves from mines when they are sent to the mine field. This method has low error rates and weather and soil features have not effected on the system[8]. The disadvantage of this method that it is slow and cannot search for landmines at depth. The system efficiency is affected by The vegetation that may interfere with the laser vibrometers that used to sense the vibrations at the soil.

2.6. Biological Methods

Mammals and insects are used to discover explosives existence in mines. Chemical methods depend on explosive components existence instead of physical properties of the subsurface. Chemical sensors have the potential used with metal detectors to reduce false alarm rates [11]. The oldest trained dogs were first landmines smelling operation to detect them.

2.6.1 Dogs and Rats

Using trained rats and dogs can smell explosives in a buried mine underground up to 60cm. Some researchers at the University of Antwerp used trained rats to detect mines [12]. The rats were trained by food rewards to sense the explosives presence by scratching the soil surface with their feet. Rats did the mission well under any field conditions. The limitation of using dogs and rats that when they are trained to detect high levels of explosives, they may not detect lower levels of explosives and may need to be trained specially for this lower level.

2.6.2 Bees

Trained bees are used to detect explosives in landmines. The advantage of this method that the bees have no false alarm. They are very fast.

2.6.3 Bacteria

By spraying bacteria on the landmine field; the bacteria will grow for several hours. Engineered Bacteria used to detect the explosives in mines. Bactria Can cover a large field in a short time. The transport of explosives on the ground will limit the performance potential of this method[11].

2.6.4 Antibodies Method

Using crystal oscillating at a certain frequency and containing antibodies to detect explosive objects. The leaked vapor from mines causes antibodies to detach themselves from the crystal and attack. The antibodies from a change in the mass of the crystal, which causes a change in the frequency. Then the operator analysis the difference in frequency to detect mines. This method can detect mines with a good rate. The limitations of this method that reacted antibodies are not reusable. The presence of reacted antibodies could also change the sensitivity of the sensor.

2.6.5 Chemical Methods

There are many researches in a chemical method such as “Dog’s Nose” program, to make an explosive odor detection device. The technique aimed to use arrays of polymer sensors that detect explosive odors. The vapor sensors use novel fluorescent polymers however like this has the most reduced identification rate. The device detects the electronics signal then the operator can decide if there are mines or not. The system detects explosive vapor from 10 o15 g concentrations per milliliter. This method can work with another detection device that rely on physical contact with the landmine field. This vapor device is small, light, portable, and simple and needs low power. The approach of this system does not perform well in very dry environments. Chemicals reactions with polymers may cause false alarms[12].

2.7 Mechanical Methods

In those methods, depend on advice that use physical properties.

2.7.1 Probes and Prodders

Prodding is most used technique in the demining process. Prodder is a stick of metal with a long 25 cm; it scans the ground. When it detects an undefined object, the operator assesses the contour, which tells whether the object is a mine or not. The deminer should have the experience to know the landmine signal sound and other underground objects. Probes decrease the risk to operators by providing them with detections. The probe can deal with acoustic, electromagnetic, thermal, chemical to increase the detecting rate. However, using probes is risky. As The operator may lose his life if any mine is triggered by prodding encounter mines that have been moved.

2.7.2 Mine Clearing Machines

When The time is very short to clear a minefield, the army will use the specific machine to pass over landmines and clear them. There are many kinds of landmine clearing machines. The machine is controlled remotely to minimize the risk to operators. There are three methods used to detect mines with clearing machines. Flailing chains to beat the ground or, rollers to roll over and trigger mines, blades to plow through the minefields, pushing the mines to the side[14]. These ways are fast and efficient and safe for operators. The limitation of this method that the area will be destroyed by citizens and The machines may miss mines.

Abbildung in dieser Leseprobe nicht enthalten

Fig 2.2: mine machine [7].

2.8. State of Discussed Solutions

Landmine sweeping and removing mines is risky, slow and costly operation. The best approach is using robot with more than one technique, through signal processing, decides whether a mine exists or not. A robot is an efficient method for mines detecting; however,not flat areas make the robot locomtion and demining more difficult. Each method has stregnths and limitations. Table (2.4) shows a comparison between the different used techniques. The main aspect for comparison are human safety, various locations accessibility, the speed of detection, and provided information from detected mines. The mobile robot with vision system is the best technique for the landmines detection.

The mobile robot gives the best the human safety as this technique depend on the vehicle on in the field of mines. The robot can detect mines with high detection speed as it uses some sensors that depend on physical features in mines. The robot can also,give the information about the nature of the landmines field and estimate accuratley the landmines existance.

To clear up landmines, two main steps used to be done. The first step is to detect mines position. The second step is to destroy the mines. The tatistics that were estimated by the Canadian Red Cross showed that a mine can cost up to $30; whie, alandmine removing can cost up to $1,000 [6]. The available solutions all over the world are interested in the defining the mines positions and then maping the existence of mines . The second step is using many different techniques for demining process. The demining method can be done by physical techniques.

Table 2.4: Mines detection solutions [6].

Abbildung in dieser Leseprobe nicht enthalten

2.9. Robot Detection Strategies

The robot can use many sensors to collect information from surrounded environment. The used sensors are such as ground Penetrating Radar(GPR), Infrared, Ultrasound and camera sensors. GPR can detect landmines where located on the ground or underground. It works better on dry surfaces; however, it cannot be used in high-conductivity areas such as soils that are salt contaminated or clay soils. Furthermore, the GPR has high-energy consumption.

Infra-Red (IR): IR is know also as thermal radiation as the reflected radiation is detected in the form of heat. IR sensors can detect underground or onground landmines. The type of soil may influence the detection as the water content in the soil may produce faulty detection. IR does not need much energy consumption unlike GPR. Ultrasound (US): The US emits ultrasound signals and collects reflected signals from the surroundings.. Camera: A camera used to detect mines that located on the ground surface, whether in humid, dry, or any other conditions. Both the camera and the US sensing techniques meet five out of the six set criteria.

The most important role in demining is to guarantee operators are in safe zone. The method of detecting mines by robots is effective because it decreases the risk, and the cost compared to manual detection [12].

Table 2.5:the Comparison of sensors [15].

Abbildung in dieser Leseprobe nicht enthalten

Robots have many advantages such as a low pressure on mine, so the probability that mine is triggered is low. Landmines detection by robots provides available low-cost sensors and high information rate.Using vision with the robot facilitates the process of collecting data in the real world in minimum time. Using improved sensors, efficient manipulators, and mobile robots help in demining efficiently [16]. While demining operators would be in a safe place, by using intermediate teleportation between operator and machine that is known as collaborative control [16]. There is a limitation in using robots in demining as some robot's structure is very expensive. Even Using inexpensive robot with simple algorithm cannot deal with obstacles problems [5]. The development of light, low cost, telerobots with an operator who monitor overall process is critical research [17].

The aim is to cover all minefield; robots should have the ability to explore to increase detection rate. A low cost demining robot must not be heavy [14]. Another area of research is Swarm robotics that combine related simple robots would be useful for mine clearance [18]. The swarm robot is simple and cheap, equipped with some sensors. The swarm is used to execute a complex task that cannot be done only by a lone robot.

2.9.1 Disadvantages of Robots

As we introduced, robots are the very effective solution for landmines detection but there are some disadvantages like Robots cannot respond to emergencies, this can cause inappropriate and wrong responses. The power requirements problems. Robots may have limited capabilities in Freedom degrees and real-time response dilation. Robots are costly.

Landmines detection process can be detected by different methods. Every method has its strengths and limitations. Using robots in mines detection is a very effective method as it provides various available effective sensors can be used to sense mines. Using robots provides safety solution to save deminers lives. Using a robot that has multi-sensor system can increase the reliability of the demining process. There are many ways can be used to detect mines with different implementations. Using light robot is more efficient than heavier one.

Chapter 3 Demining robot techniques

There are many techniques that can be used by robots to detect the landmines.The motion planning and Sensor fusion are very important techniques which define the accuracy of demining process.The motion panning problem is very essential part in the robots which can be used to detect mines. The motion is very important also in collecting available information with sensors about the environment.The Sensor fusion is the way that combines data from different sources to collect all available information to help in decision making .There are different techniques as levels are used to fuse data from multi-sensors.For example ,low level that deal with signals at a low level to collect information from the different sources[18].

This chapter introduces the motion planning techniques which can be used by autonomous robot to find the optimal solution in an environment.!: discusses how the path planner problem can use the environment information to facilitate the motion decision.It give overall idea about motion types and how the robot can simulate biological system to locomate.It introduces some techniques of the motion techniques and when they can be used.It also ,describes how the motion planningcan be used to form a map from the environment with available information.It also presents the sensor fusion concept and the process of combining information from different sensors. It also ,present the sensors fusion importance and the reilablity of integrated data.

[...]