What other tests or uses are scheduled for the robot in the near future?

In April, Crabster has deployed to the site of Sewol ferry disaster to assist rescue operations. But originally, Crabster was supposed to explore ancient shipwreck in west sea of Korea during that time. So, we are planning again to explore the ancient shipwreck in next spring.

What are the key mechanisms that make Crabster work?

Crabster is the robot for underwater exploration. So the most important key technology is the water proof and pressure resistant mechanism. Crabster is designed to be operated in 200 meters depth so it has to endure minimum 20 bar of the water pressure. The second is the harmonious and stable control of many motors in six legs. Crabster has six legs and thirty BLDC motors in the legs. To walk on the seafloor or explore underwater, Crasbster has to be control the motors harmoniously and stably. All joints are connected to the computer via CAN bus. The computer computes commands for the harmonious control of all joints by solving kinematics and dynamics of all the legs. Crabster can walk with various gait algorithms to explore on the seafloor. Crabster can perform posture control to survey or work on the seafloor.

What were the chief challenges to designing Crabster?

One of the challenges was to design a hydrodynamically streamlined body. The body shape has to minimize the fluid drag while securing sufficient space for the exploring equipment. The body also has to be lightweight but structurally strong. We decided to adopt composite material to satisfy the competing demands. We designed the frame using carbon fiber reinforced plastic and the skin using glass fiber reinforced plastic.

Another challenge was to design an underwater joint mechanism. The joints of legs have to be powerful, no backlash, no exposure of wiring, watertight and compact. We adopted frameless BLDC motors and harmonic drives for the joint mechanism. More than ten thousands parts were used in a front leg and the leg has more than 60 O-rings for watertight.

Will Crabster help out at the ferry site in the future?

I hope so, but I don’t know. We reported our survey results to government and government will make decision.

Was it ever difficult to coordinate Crabster's work along with all the other ships, robots and human divers operating at the ferry site?

Hundreds of ships were float around the ferry site and dozens of boats were float around the barge on the ferry site. It was not easy to work along with other ships because the velocity and direction of current were changed continuously. During the Sewol survey, our ship drifted to the barge and Crabster’s tether cable touched the anchor line of barge. We had to have help from a tug boat to escape the situation.

How did Crabster compare with the robotic submersibles (AUVs and ROVs) that were deployed at the ferry site?

Because Crabster weighs 150~ 200 kg in water, it can keep its posture and position more stably on seafloor even in current. Crabster has acoustic instrument against turbid water. In addition to the default sonars, we installed additional sonars when we were waiting in pier. We had additional scanning sonar for vertical scan and additional multi-beam sonar for real-time image in more range. Our team developed AUVs and ROVs for last 20 years, those propeller-driven subs tends to drift by current which is resulted to entangle the tether cable with mooring line of other ships or barge. I think the ROVs deployed at ferry site were not different.

Did Crabster provide any advantages for the search and salvage operation over human divers? Was it able to work together with human diver teams?

Since the visibility at the site was only 0.2 meters, divers were hard to search underwater. Crabster can stay deeper and longer, Crabster can see farther. But Crabster cannot go into the ship. We wanted to work together with human divers, but we had no chance to do.

Were there any days when the water conditions were too rough for Crabster to deploy? Or was the main challenge bad weather that prevented the mothership from deploying Crabster

We were in Jindo for 20 days. We surveyed Sewol ferry for 2 days. And we waited at pier for 15 days. Since there were no requests for additional survey from government, we decided to survey seafloor at a distance from Sewol for ourselves for 3 days. But we could not deploy Crabster at first two days due to the bad weather conditions. The weather was windy and wave height was 2 meters, the rolling of our ship was too big to deploy Crabster. The last day we sent Crabster to seafloor and we found that the 6m-pipe- debris, which were detected by side scan sonar before a couple of days, were not from Sewol.

How much time did you usually have to operate Crabster underwater each day? Were there any time limits based on Crabster's power supply, outside operating conditions or official limits set by the government officials directing the operation?

When we surveyed the Sewol, the coast guard gave us just one and half hours, but we surveyed more than two and half hours. When we surveyed 0.5~1km distance from Sewol, there was no time restriction to survey except mothership’s drift. Crabster remained seafloor until the mothership drifted apart from Crabster by wind and current. Crabster used to stay on seafloor 0.5~ 2 hours at a time.

How much time did it take to deploy Crabster each day by putting the robot into the water?

It took 5 minutes to deploy Crabster into water and another 10 minutes to touch bottom 45 meters below surface.

Were there any difficulties in the mothership putting Crabster into the water with its crane and retrieving Crabster from the water? And did Crabster walk and operate well underwater?

We used our own LARS(launch and recovery system) consisting of crane and cable winch. It was the first use of our LARS designed for Crabster. The LARS worked well. During the weather was good, we have no problem with mothership and all the launch and recovery process was done perfectly. But the ship was too small to work in wavy and windy weather. The ship’s roll motion was too big to launch and retrieve the 700kg-Crabster safely. Crabster walked seabed while taking seafloor images with acoustic camera. Crabster was launched 13 times and stayed in water total 15 hours 36 minutes and worked total 11 hours 21 minutes on seafloor while Crabster remained in Jindo for 20 days.

What parts of the Sewol ferry did Crabster inspect? Were there any areas that Crabster had to avoid because of dangerous conditions or other factors?

Crabster was deployed 70 meters away from Sewol. Crabster could see the contour of bow master and upper cabin of Sewol. We wanted to go more close to Sewol but we could not have permission due to the safety of divers. We were very cautious so that Crabster’s tether cable was not entangled to the anchor line of barge. If we could attach our ship to the large barge, we would get more information about Sewol and her surroundings.

What were the most crucial instruments that Crabster used during its deployment at the ferry site? What kind of information did Crabster collect or find during the operation?

Most crucial tool of Crabster in the site was acoustic instrument on the six-legged platform. We could see nothing with ten optical cameras on Crabster. We got acoustic images of sunken Sewol from scanning sonar. We got acoustic images of seabed from acoustic camera. The six-legged Crabster worked well as a stable moving station for the acoustic instrument in tidal current. The Sewol ferry was slanted 90 degrees to port side. The seabed was composed of mud, stones and small pieces of shell. Crabster also acquired depth, altitude, conductivity, temperature. Attitude of body, all joint angles and electric current of six legs were also monitored and recorded in surface control room.

Can you describe the challenging water conditions that Crabster had to navigate? For example, how fast were the currents you had to deal with? How deep was the water? How was the visibility underwater?

The water depth was about 45 meters, visibility was less than 20 centimeters and water temperature was 11~12 degrees C. The maximum current is more than 8 knots at the area but when we were there, the tidal current was less than 3 knots. Sometimes tough wave and wind bother us. The Crabster had no problem to work in those conditions, but Crabster cannot overcome its mother ship. Our ship was too small to keep her position in those current, wind and wave conditions. So we wanted to attach our ship to the large rescue barge firmly moored with four anchors. But we could not have permission.

What were the main goals in using Crabster during the search and salvage operation? What was Crabster's main role in helping assess the ferry's condition or helping to search for people inside?

The Sewol sank at 45m water depth and the water was very turbid. It was dangerous for human divers to search the seafloor. I thought we could inspect the hull of sunken ferry and debris around her using Crabster’s acoustic instrument. And the information would be helpful for the planning of rescue and searching activity. When we reached the ferry site with Crabster system, the divers were going into the sunken ferry continuously for rescue. The highest priority was on the rescue activity by divers and all other team was restricted to approach to the rescue barge. Since Crabster is not able to go into sunken ferry, we also had to wait for approaching permission keeping apart from the rescue barge.

Did you and your colleagues volunteer Crabster to help out at the ferry site? Or did government officials request your help?

When I knew the rescue team had serious difficulties due to the high current and turbid water at the accident area. I called Dr. Suh direct general of KRISO and talked about Crabster’s functions and possibilities for the helping of rescue. Task force team of my institute reviewed underwater robots made in KRISO and agreed to send Crabster to the area. The decision was not easy because Crabster was still under development and the performance was not fully verified. Two days later, government requested our help for surveying of sunken Sewol. We moved the Crabster system from Namyangju to Jindo next day, 21 April.

When your full size Crabster is completely built, what will be the first body of water where it is tested?

We will finish the integration of CR200 at the end of April. The hardware, software, communication and various algorithms will be integrated and tested in April. From the May, we will conduct the integrated shop test for CR200. We will check the integrated functions of CR200 such as remote control of posture, walking and interface of various sensors in May. If there is no problem in the shop test, we will go for sea trials in mid of May. The first sea trials will be conducted at pier of Geoje branch of KIOST which is located in the south of Korea. As you know, however, the schedule is changeable because there may be unexpected situation.

With the 1/2 size prototype - where can it walk, what exactly can it do in order to demonstrate? Is there something fun or visual that you have experimented with?

Little Crabster(LCR200) can change its posture using six legs, and can walk to any direction like crab. LCR200 can react to external disturbance for balancing by center-of-pressure-control. The shape of LCR200 is different from CR200 but the kinematic structure is the same. We use LCR200 for the pre-test of control algorithms of Crabster

How do you expect marine life do react to it – will animals treat it as a crab?

I don’t know if the animals treat the Crabster as a real crab or not. But we observed many fishes during the first underwater test. I thought fishes like Crabster more than the other underwater vehicles with propellers. Crabster has less flow and noise than propeller-typed vehicles. I hope animals treat Crabster friendly.

How could the concept be developed further? What new solutions could it potentially provide?

We are now studying to make the Crabster to swim like turtles or diving beetles. And we are also considering the hydraulic Crabster for heavy duty underwater working.

How much demand is there – do you expect Crabsters to be manufactured for use across the world?

Crabster CR200 will be used as a platform to develop and test the new technologies for underwater working. Now, I don’t know how much demand will be. But I expect there will be demands in scientific fields if Crabster does well in its sea trials. Then many other Crabster will be possibly manufactured for use across the world.

What specific assignments do you expect it to be used for and what is a realistic timescale?

Artifacts have been found at more than 250 points under the West Sea of Korea. However, only 20 of them have been excavated due to the harsh environment of current, low temperature, high turbid and depth. Though there have been many trials with ROVs for exploration in the area, underwater archaeology specialists never get results due to the high current and high turbidity.

Crabster's mission will be seafloor exploration for scientific research such as marine geology, biology, and archeology. We suppose CR200 conducts seabed mapping, survey and inspection of wreck, pipeline, scouring, waste, ecosystem and pollution in off the coast down to 200m depth. CR200 will help divers or work instead of them in harsh environment with high current, deep, turbid and low temperature. It also could assist in locating underwater resources, carrying out underwater mining, responding to oil spill incidents

We will conduct a tank test with CR200 in the first week of April (next week) in Daejeon. We will check the current-enduring ability of CR200. In May 2014, we will go to West Sea of Korea for the first mission of CR200. The mission will be excavation of ancient artifacts by cooperating with marine archaeologists. CR200 will do precise seafloor mapping of the turbid area with acoustic equipment. We will also try a number of possibilities of CR200’s role for underwater excavation.

How does the crab-shape design make it better suited to this exploration – what can Crabster do that other designs/crafts cannot?

Some lobsters live in stream and some crabs live in coast where tidal wave exists. They can live there because they constantly control their posture. If they don't control their body like dead crab or lobster, the current will blow them off. CR200 mimics their shape and behavior.

A.     CR200 has advantages over ROVs/AUVs in strong tidal current as follows.

 i.         CR200 can endure in high tidal current environment better than AUVs or ROVs because CR200 walks on boundary layer of seafloor where current speed is relatively less than upper layer.

 ii.         CR200 can keep or adjust its posture stably based on the mechanical contacts between legs and seafloor. CR200 is able to improve its tumble stability by move its footing.

 iii.         CR200 can stabilize its posture by utilizing the hydro forces acting on the body and legs. CR200 can optimize its attitude to reduce the drag force and increase the downward lift force to improve the tumble stability.

B.     CR200 has advantage to overcome high turbid underwater environment

 i.         In the very turbid area, underwater robot has to use sonars such as scanning sonars or acoustic cameras. CR200 provides very stable moving base for the acoustic image equipment.

 ii.         Moreover, moving underwater robots in turbid area may be very dangerous because there may be obstacles around the robot like abandoned fish nets and ropes which entangle the robots. CR200 is safer in those areas due to the stable walking without drift.

C.     It is very dangerous and limited for human divers to go to following area

  i.         deeper than 30m

  ii.         temperature is less than 10 degree C

  iii.         current is more than 1 knot

  iv.         visibility is too short

What challenges do you still face?

The challenges we face are 1) to enhance the moving speed of Crabster (presently Crabster’s walking speed is 0.1 m/s), 2) to enhance the posture and walking stability in strong tidal current 3) to enhance the walking stability on more rough terrain and 4) to handle the drag force acting on tether cable in strong current.

What development may be required for it to be used in deep-sea exploration?

Crabster requires for deep-sea exploration 1) the watertight multi-joint leg mechanism, 2) control algorithms of six legs for underwater walking and manipulation, 3) posture adaptation for current and 4) underwater vision system for high turbid and low visibility environment

How has the Crabster performed in tests?

The first underwater test was conducted from 24 June to 10 July 2013 at pier in South Sea Research Institute of KIOST which is located in the Geoje City.

In the test, we successfully checked and verified 1) the watertight of dynamic seals of thirty joints, 2) identification of topography and underwater structures with scanning sonar and analyze the effect of posture changing on the scanning image, 3) real-time identification of objects with acoustic and optical cameras, 4) underwater position tracking of CR200, 5) gathering the scientific data, 6) underwater walking and posture control with six-legs.

And we are performing the on-land shop tests nearly every day. We upgrade Crabster software for more stable and fast walking and manipulation. We revise the hardware mechanisms for more stable and robust functions

The Crabster did very well in the tests until now. It is very exciting to see a newly developed creature works as our will.

What kind of jobs can Crabster do?

We suppose Crabster can conduct seabed mapping, survey and inspection of wrecks, pipelines, ecosystems and pollution down to a 200-meter depth. Crabster will help divers or work instead of them in harsh environments. It also could assist in locating underwater resources, carrying out underwater mining, and responding to oil spill incident

How can Crabster go where scuba divers can't go?

Scuba divers are recommended to dive within 30 meters water depth, less than one knot current, and higher than 10 degrees C temperature. But Crabster is free from the safety problem because it is robot. Crabster was designed to dive up to 200 meters water depth and to endure 3 knots current.