2003 Safety, Security and Rescue Robotics Workshop (SSRR)
 |
 |
|
|
In the United States, only certain rescuers can enter an Urban Search and Rescue (USAR) environment, which usually consists of collapsed buildings resulting in large rubble piles, toxic gases and even possibilities of secondary collapses, which we saw at the World Trade Center (WTC). On top of having regular fire rescue credentials, one must also possess special certification to be fully qualified for USAR. The number of qualified personnel is therefore limited. Robots would facilitate a search and rescue effort of this magnitude tremendously by helping out in four crucial areas:
The goal of the 2003 SSRR workshop is to set a research agenda and vision for robotic USAR, as well as form a scientific community. The SSRR workshop will be dedicated to identifying and elaborating on the key scientific issues which are Mobility, Sensors, Perception, Human-robot Interaction, and Distributed Intelligence and Communications.
The National Urban Search and Rescue Response System, established under the authority of the Federal Emergency Management Agency (FEMA) in 1989, is a framework for structuring local emergency services personnel into integrated disaster response task forces. These task forces, complete with necessary tools and equipment, and required skills and techniques, can be deployed by FEMA for the rescue of victims of structural collapse. Sam Stover is affiliated with Indiana's Task Force 1.
All FEMA teams enter a search and rescue situation comprised of the personnel shown in Figure 2. When the team first arrives on scene, an initial survey is done where such things as safety factor, probability of survivors, probability of locating survivors, construction type (e.g. wood, steel, etc.), and site conditions (wind, temperature, etc.) are assessed.
The equipment used to support a task force weighs nearly 60,000 pounds and is worth about $1.4 million. This equipment includes things like an electric hammer drill, electric viewing equipment, electronic listening devices, and structural evaluation devices. There are different types of cameras that are traditionally used by tasks forces nationwide, however, they all have their drawbacks. The Searchcam is heavy and is not flexible or waterproof. Infrared and thermal imaging cameras are generally used to see through rubble, but a body will only give off enough heat to be seen by one of these cameras for the first 48 hours of the search - rendering them useless after that. Life detectors, dubbed the best device by many teams, is an electronic listening device that uses triangulation to pinpoint a victim's position. When operating this device, it is standard FEMA protocol to cease all ambient noise. However, this did not happen in the WTC rescue effort because there were several volunteers helping out that were not qualified to be there.
Canines play a very important role in the search and rescue effort because of their keen sensing abilities. There are however, pros and cons associated with every canine. Canines can search a large area in the shortest amount of time, work in dangerous areas, detect unconcious victims and gain access to hidden areas. However, disadvantages always come with the advantages. When a canine senses something it will usually bark to notify the canine speciallist. However, it would be considered a waste of time, which is very limited, if the canine sends the specialist on a wild goose chase. Therefore, another canine is needed to confirm barking. Since there are only four dogs per team, this puts the task force at a large disadvantage. Canines are also not invincible - they require rest periods. A typical canine can only search for a couple hours before becoming tired and dehydrated. Furthermore, canines cannot be used in a biological or chemical attack.
There are ubiqitous problems in search and rescue efforts that hinder the efforts of Task Forces. When a victim is buried and unconcious in a pile of rubble, it is impossible to assess his/her vital signs. Furthermore, the most critical parameter in a search and rescue effort being time, assembling a task force team can take hours. USAR teams are made up of civilians and firefighters from different backgrounds and it usually takes 2 to 4 hours to gather everyone up. In most disasters that involve a collapsed building, the largest obstruction to a rescue effort is the "widow maker". A widow maker is usually a beam or structure that is on the verge of collapsing and FEMA task force members (including canines) will not step foot anywhere around the area. WTC Task Force members named the beam in Figure 3 a widow maker.
The problems and drawbacks associated with the equipment, canines and mother nature motivate the need for robots in a search and rescue environment. Robots can fit in many places where humans and canines cannot. Even if the robot is too large to fit in a rubble opening, it will usually have some sort of extendable snake arm with a camera on the end to obtain a close-up view of the victim. The most demanding place for robots is in a Weapons of Mass Destruction (WMD) attack. The chemical suits that must be worn by all personnel in this situation are bulky and cumbersome and will limit mobility. Robots, of course are not sensitive to biological and chemical agents and can be deployed while command and control teams sit safely inside a communications center. Robots are needed to create an expansion of senses for task force teams that include vision, smell and sound. But most importantly for FEMA personnel, something is needed that is backpackable, able to cover a grid of 1 city block, is equipped with lights for power outtages and light enough for a human to carry.
The mission of SPAWAR is to provide the warfighter with knowledge superiority by developing, delivering, and maintaining effective, capable and integrated command, control, communications, computer, intelligence and surveillance systems. SPAWAR provides information technology and space systems for today's Navy and Defense Department activities while planning and designing for the future.
The Mobile Detection Assessment and Response System (MDARS) is a joint Army-Navy development effort to provide an automated intrusion detection and inventory assessment capability for use in DoD warehouses and storage sites. The goal of MDARs is to provide multiple mobile platforms that perform random patrols within assigned areas of warehouses and storage sites. The patrolling platforms can detect intruders and anomalous conditions such as flooding or fires and determine the status of inventoried items through the use of specialized RF transponder tags.
SPAWAR is looking into a lot of new areas such as Unmanned Surface Vehicles (USV) that can detect mines, expand NAVY SEAL capabilities and increase harbor security. They also are anticipating future projects in which Unmanned Aerial Vehicles (UAV) and ground-based robots work in unison.
Gunnar Kuepper is Chief of Operations with Emergency & Disaster Management, Inc., in Los Angeles, CA. This independent agency advises private, non-profit, and governmental organizations throughout the world in comprehensive emergency/crisis management and business continuity programs. Much of Gunnar's planning and consulting work takes place within the international aviation industry. He and his team at EDM have analyzed the emergency management procedures and response operations in numerous catastrophic incidents, ranging from transportation accidents, to fires and explosions, to acts of terrorism and ultraviolence.
There are a lot of events that can take place in an airport that have catastrophic consequences. A fire in an airport terminal would be deadly. Although the terminals have hundreds of would-be escape windows, they are all unbreakable for security reasons. This makes ingress and egress routes very limited. Furthermore, airports are the center of gravity when it comes to recent terrorist attacks. Terroritst attacks have either occurred or initiated at airports. And since 70% of all terrorist attacks involve explosives, hand grenades or bombs, airports are very velnerable to emergency and disaster environments.
With situations like this, the use of robots in airports is critical. Robots could be used in airports for such tasks as chemical and biological agent detection, evacuation of victims, and surveying a bomb threat area. Airports receive several bomb threats per month and in most cases the airport shuts down all operations until they determine the extent of the threat. This costs the airport a lot of unnecessary time and money. However, sending in a fleet of robots to detect, pinpoint and assess the presence of any live bombs (or false positves) would be invaluable. An aerial robot fleet, in particular, possesses the capability of flying over humans and would be less intrusive than a ground-based robot fleet. Once the threat is then confirmed (or discarded), evacuation procedures can then begin.
Radiance Technologies applies emerging technologies to improve military capabilities and matches technology solutions to meet military needs. Dr. Frank Rose, from Radiance Technologies gave a presentation on the future of the army's unmanned ground vehicle technology.
By the year 2013, one-third of all military vehicles will be unmanned. It is hoped that these unamanned vehicles will be 100% autonomous. The word autonomous can be interpreted in different ways because of the different levels of autonomy (see Figure 5), but the military is expecting fully autonomous operation by 2013.
There is several key factors to take into consideration when designing an autonomous robot. They are perception, planning, behavior skills, navigation and learning and adaptation. Dr. Rose recommends giving the highest development priority to perception. The many different platforms of autonomous robots the military is working on include TGV (Teleoperated Ground Vehicle), SAP/F (Donkey), PC-AGV (Wingman), and Hunter-Killer.
Robin Murphy is the pioneer in the rescue robotics industry and heads the Center for Robot-Assisted Search and Rescue (CRASAR) at the University of South Florida. CRASAR deploys on order into crisis sites and hazardous areas around the globe with self sufficient robotic systems that can operate in areas that are physically inaccessible or operationally undesirable for humans, canines, or other search and rescue assets. Their specific objective is to augment and support existing rescue organizations, not replace them.
The research gaps that plague the rescue robotics industry will hopefully be overcome in the near future. Vertical drops are obstacles that are frequently encountered and can be as much as 10 or 20 feet and certainly would destroy most robots. Wireless constraints include 802.11 saturation (just about everyone has something running on 802.11) and wireless dropout. There are also visual limitations. You cannot tell if an unconscious victim is dead or alive by looking at a noisy image. In the WTC search and rescue effort, Robin claimed that 12% of the time was spent tasking and the rest was spent on trying to interpret what you were looking at. There were several cases where the robot passed right by a victims. You also need to determine if the victim is dead or alive. Endurance is also shortchanged for typical 12 hour search and rescue missions; the average mean time before failure was approximately 8 hours. Another setback is the robots are extremely labor intensive usually taking 2 or more people to operate 1 robot; 1 person to drive and the other to "look". Additionally, the interfaces are not user-friendly.
To overcome these gaps in SAR missions, the robots need to be man-portable, controllable while wearing protective equipment (chemical suits), waterproof, substitution supportive (both hardware and software), and interfaced to a computer for image processing. They also have to be equipped with color video and 2 way radio capabilities, built-in mechanisms for recording and playback of sensor data, proprioceptive sensors (can tell pose, shape, etc.), adaptive bandwidth devices and user-friendly interfaces. And finally the endurance standards have to be increased (12 hr shift with intermittent use). The robot platforms must include elephant trunks and snakes to extend cameras in hard to reach places or tether management robots (marsupials) to release a smaller robot if the situation calls for it. They must contain deployable payloads so the robot doesn't have to remain with victim and have interchangeable parts (software and hardware).