Search and rescue (SAR) operations pose grave risk to human lives, both to victims and those sent in to save them. More than 700,000 people were killed as a result of disasters worldwide between 2005 and 2014, according to the United Nations Office for Disaster Risk Reduction (UNISDR).
Robots are being developed and increasingly used in SAR missions to help cut costs and improve efficiency. Timing is everything when lives are on the line, and these machines can not only reduce the duration of these missions, but also arm SAR teams with an arsenal of information, which helps them arrive on-site quicker.
Some of the areas where robots have been implemented to aid in SAR efforts:
- Mapping and maneuvering of disaster areas after natural calamities, terror attacks, accidents, explosions, etc.
- Robots are also providing assistance to firefighters worldwide
- Bomb squads are using AI to aid in defusal and disposal missions
- AI is being developed to detect signs of life, such as heart beats and breathing, of victims trapped in the aftermath of natural disasters
- Drones are being used to distribute water, food, medicines and other supplies
We’ll cover each of these five application areas in depth in the sections of the article below:
Disaster Area Mapping and Maneuvering
SAR teams are tasked with exploring extensive areas in a short period of time to find those trapped or stranded after disasters. Time is of the essence in these kinds of missions, and state-of-the-art robots are employed to map out areas that have been destroyed, giving rescuers an edge and improving the efficiency of their operations.
Unmanned Aerial, Ground and Marine Vehicles
The Center for Robot-Assisted Search and Rescue (CRASAR), which has been researching and developing technology for disaster use for several years, has sent in many unmanned aerial, ground and marine vehicles to aid in SAR missions, including during Hurricane Katrina, and these machines are also now used in Italy and Germany. Human-robot teams, for both unmanned ground vehicles (UGVs) and unmanned aerial vehicles (UAVs), have shown a nine times increase in performance as opposed to human-only teams, according to the CRASAR.
A video of some of the search and rescue operations that CRASAR has been a part of between 2001 and 2016:
PrecisionMapper
Commercial drone and data company PrecisionHawk has developed a host of UAVs, including the Lancaster fixed wing drone, and data mapping technology such as PrecisionMapper, which processes aerial data into 2D or 3D products. It has an expanding library of analysis tools, which makes simplifies data sharing. The company’s Orthomosaic algorithm can create several surveys in no time and convert them into a PDF with images of various compass directions, and information on latitude/longitude, map projection, resolution and weather.
An image of the PrecisionHawk drone at scale
This company video demonstrates how the UAV collects and exports visual and sensory data: https://youtu.be/-TLl_hvDIQk
The Algorithm Marketplace, launched in 2015, provides automated analysis of aerial data acquired by UAVs. Using sophisticated algorithms, drone operators can interpret aerial data with an action-based report, which can improve management of assets. These algorithms have already been used in agriculture, but can also potentially be used in SAR missions.
PrecisionHawk explains how the Algorithm Marketplace works:
RoboSimian
The ape-like bot developed by NASA’s Jet Propulsion Laboratory (JPL) has sensors in the wrist and ankle that enable it to gauge terrain as it moves. This non-humanoid machine has four limbs, capable of both mobility and manipulation, to achieve passively stable stances; establish multi-point anchored connections to supports such as ladders, railings and stair treads; and brace itself during forceful manipulation operations.
JPL is testing the robot, which is controlled by humans, for wider application in new, complex environments, from deep waters to outer space, according to a report from NASA Tech Briefs.
Tested interviewed Katie Byl of the UC Santa Barbara Robotics Lab and she explains how RoboSimian functions:
Fire-Fighting Bots
Firefighters risk their lives every time they put on their suits and jump into a fire.
There were nearly 30,000 fire-fighting-related injuries in 2015 in the United States, according to the National Fire Protection Association. With advances in technology, these brave men and women can get some much-needed help from machines that can withstand higher temperatures, move with agility and are unaffected by thick smoke or heavy flames.
THOR
THOR, the Tactical Hazardous Operations Robot, is the first bipedal in the world to extinguish a fire at the final challenge of Phase I of the SAFFiR program. The robot, which was designed by a team of Virginia Tech scientists, is a 33 degree of freedom robot capable of compliant, force-controlled walking. A team of TREC, Terrestrial Robotics Engineering and Controls Lab, engineers designed the machine in-house.
The humanoid robot has the ability to directly control its ground-reaction forces, enabling it to walk across concrete, grass, gravel, and even sand, with its 3D walking feature that is achieved by analyzing inverted pendulum dynamics and stabilizing the center of pressure by applying ankle torques on the ground.
The creators of THOR explain how the robot works:
Thermite
Fire-fighting bots such as Thermite, developed by Howe and Howe Tech, are remote-controlled machines with multi-directional nozzles backed by pumps that deliver up to 600 gallons of water per minute. These robots help first respondents by reducing the flames, enabling firefighters to get reach otherwise hostile environments after industrial, nuclear and chemical fires.
Watch the Thermite in action here:
Bomb Disposal AI
Machines used in bomb squads can ensure for a safe and accident-free bomb defusal or disposal mission. These are very tense and critical situations that could cause unnecessary loss of life. Drones can go in and save the day with little risk of an accident. Usually, the drones work with a remotely located bomb squad that monitors surroundings, analyzes data sent in and helps machines with bomb diffusal/disposal.
tEORor
Telerob’s remote-controlled telerob Explosive Ordnance Disposal and observation robot, or tEORor, provides enhanced capabilities to explosive ordnance disposal (EOD) teams. The robot can disarm traps, fireworks, improvised explosive devices (IEDs) and other hazardous materials, with high reliability and top-notch movement. This bomb disposal robot system is currently servicing military and law enforcement units in more than 41 countries.
T7
The UK Ministry of Defence (MoD), in a deal worth up to £55 million, in September 2017 contracted U.S. robotics company Harris Corp to supply 56 T7 robots as part of the MoD's Project Starter. The T7 robot can be used in various areas including in commercial and military missions such as hazardous materials (HAZMAT) cleanup, explosive ordnance disposal (EOD), intelligence, surveillance and reconnaissance (ISR) and special weapons and tactics (SWAT) missions.
The T7 EOD robot’s manipulator is controlled using a haptic controller from a distance of up to 1km. The controller incorporates daylight visible display, two-way audio communications system, and 4-DOF (degree of freedom) haptic feedback function.
The haptic controller system enables the operator to create preset sequences of tasks for the robot manipulator. It can perform intuitive spatial control with pistol grip over the actuated object. It can also operate the manipulator in motion / rewind modes.
Harris Corp’s unveils the T7 Robot for the first time at AUSA (Association of the US Army) convention in 2016:
Studies are ongoing to develop more function-specific robots, such as robots for sniffing out bombs and disposal bots to disarm them, according to a BBC report, which added that the key to bomb disposal is to render the EOD inert without causing detonation.
Detection of Casualties
When disaster strikes it is often critical to find victims as soon as possible, with every passing minute the chance of survival reduces. Sometimes, survivors are left unconscious under layers of rubble and slowly fade away as SAR teams try to locate them. Machines have helped detect signs of life much faster hence saved many more lives.
FINDER
AI technology such as FINDER, or Finding Individuals for Disaster and Emergency Response, can detect survivors’ heartbeats and breathing under layers of rubble. Developed by JPL and the Department of Homeland Security's Washington Science and Technology Directorate, it has been deployed all over the world, including the Mexico earthquake in September 2017, in Puerto Rico after Hurricane Maria and post the earthquake that struck Nepal in 2015.
FINDER in action in Nepal:
MEOSAR
The International Cospas-Sarsat Programme, first deployed in 1982, helps in SAR distress alert detection and information distribution system to find and locate emergency beacons activated by aircraft, ships and backcountry hikers in distress. The programme’s newest technology, MEOSAR (or Medium Earth Orbit Search and Rescue), is expected to be fully functional in three to five years and could provide great assistance in locating those stranded in remote locations.
Orolia’s McMurdo in May said it would deploy its Search and Rescue Satellite Aided Tracking (SARSAT) system to a leading Southeast Asia search and rescue authority. The technology will provide access to the MEOSAR satellite system, improving accuracy and response times for SAR teams in the region, according to the company.
This video explains how a MEOSAR system works:
Relief Operations
People stranded after an earthquake or hurricane or who are living in a warzone are often stuck for days without any food, water or medicines. They are usually cut off from the world due to collapsed infrastructure, making it hard for them to receive basic necessities. AI can help deliver vital supplies to these victims much faster than most first respondent teams.
NVIDIA’s Jetson-Based Drones
Drones equipped with NVIDIA’s Jetson computing platform can be used to automate SAR missions. The software uses deep learning for object detection, path planning and collision avoidance. With a clear path and swift movement, the waterproof drone can deliver humanitarian aid to those stranded after disasters.
Snake bots
Robots, such as snake bots, can maneuver through difficult terrain looking for survivors, collecting data and also providing supplies like water and food.
Carnegie Mellon University’s (CMU) multi-jointed snake bot was deployed to look for survivors trapped in a Mexico City building that collapsed after the Sept. 19 earthquake. The bot, developed by Biorobotics Lab, provided rescue workers with video feeds of two passes through the rubble, but did not find any survivors, Matt Travers, systems scientist in CMU’s Robotics Institute said.
Researchers at Stanford University have created a pneumatic air pressure snake robot, which can grow across great lengths without moving its entire body. In future, the vine-like robot may also grow using liquid, which could help deliver water to people trapped in tight spaces or to put out fires, according to a research report.
This video shows how the soft robot moves like a serpent and can help deliver supplies to difficult-to-reach locations:
Biomimicry should continue to be a useful inspiration for search and rescue robots in the future. From robotic lobsters to UC Berkley’s robotic cockroaches, academic efforts to model the dexterity of animals continue at universities around the world.
UAVs Supply Aid
The Government of Malawi and UNICEF in July said they would test the use of UAVs in humanitarian aid distribution, including in the delivery of emergency medical supplies, vaccines and samples for lab diagnosis. The corridor is the first in Africa and one of the first globally with a focus on humanitarian and development use. It will focus on three main areas: imagery, connectivity and transport.
This World Economic Forum video shows how drones deliver humanitarian aid in Africa:
The Economic Impact of Robotic Applications for Search and Rescue
Over the past few months, there have been a series of hurricanes, earthquakes and other naturals disasters, which have impacted pretty much every country in the world.
While the full extent of the damages on economies and governments is yet to be determined, one thing is for sure with the help of technology these costs could be drastically reduced.
Debates continue to rage on the relation between robotic applications and job automation, but there’s likely to be less push-back against robotic applications in search and rescue. The extremely dangerous nature of the work make it (along with military robotic applications) an excellent “low hanging fruit” testbed for robotic capabilities.
There are a great deal of AI offerings which would reduce the amount spent on disaster relief and could also decrease the time spent on these operations, hence saving more lives and reducing overall damage.
According to the United Nations the International Community is about $15 billion short every year on the amount of money it needs to spend to provide critical support to people affected by natural calamities and wars. Several companies have begun investing in AI to bridge this gap and bring those devastated by these disasters the relief they so badly need.
More than 60 percent of people who took a survey conducted by Swiss Foundation for Mine Action (FSD) said they favored the use of machines in humanitarian work.
Robotics has come a long way in aiding SAR teams and there is only more potential for AI in the field, with several avenues yet to be explored. Scientists are researching and developing new technology and bringing software and hardware up to speed to help save lives in the aftermath of disasters.
This article was researched and edited entirely by the TechEmergence team.