AI summary by TwinMind

The Robots That Swim Like Fish

Asianometry

16m 22s

Introduction and Discovery

  • Author discovered robot fish in kiddie pool at New Taipei Tech University
  • Attempted to contact research team twice for interview without response
  • Video focus: robots that swim like fish vs traditional propeller-based underwater vehicles

Current Underwater Robot Limitations

  • Most unmanned underwater robots use rotary propellers
  • Propellers are loud and inefficient
  • Efficiency measured by thrust × forward velocity ÷ power used
  • Propeller efficiency: 40-70%
  • Fish tail efficiency: over 86% when optimized
  • Benefits: smaller batteries, no tethered power needed
  • Navigation issues: tight spaces require thrusters all over body

Fish Swimming Evolution and Classification

  • Over 500 million years of fish swimming evolution
  • First formal study: 1926 by C.M. Breder (New York Aquarium)
  • "The Locomotion of Fishes" article proposed two categories:

Body and Caudal Fin Locomotion (BCF)

  • Fish bends body and tail back/forth
  • Creates backwards propulsive wave
  • Examples: tuna, sharks
  • Suitable for long-term, high-speed swimming

Median and Paired Fin Locomotion (MPF)

  • Uses fins near head/belly
  • Example: puffer fish with pectoral fins behind gills
  • Better for maneuverability

Gray's Paradox (1936)

  • James Gray studied dolphin swimming dynamics
  • Chinese word for dolphin: "hai tun" (sea pig)
  • Calculated dolphin needed 7× strength it appeared to have
  • Traveled ~20 mph in water
  • Scientific consensus: Gray made calculation errors regarding:
    • Musculature
    • Drag forces
    • Drafting use
  • Triggered extensive aquatic animal research

Theoretical Developments (1960s-1970s)

Light Hill's Equations (1960)

  • First mathematical description of fish movement
  • Two actions: body undulation + tail fin pitching
  • Initially described small amplitude movements only

Expanded Theories (1971)

  • Large amplitude elongated body theory
  • Combined into generalized elongated body theory
  • Reactive theory based on Newton's third law
  • Most well-known theoretical description despite simplifications

Additional Theories

  • J.Y. Chang's wave plate theory
  • Attempted to improve understanding of fish swimming motions

Early Robot Development

Triantafyllou Brothers (Late 1980s)

  • Focused on water vortices from fish tail fins
  • Fish acceleration process:
    • Hard first flap creates large vortex
    • Second flap creates opposite vortex
    • Vortices combine for strong forward thrust
  • Trout/salmon exploit rock vortices in rivers

RoboTuna (1995)

  • First acknowledged fish robot
  • Body: pulleys, links, aluminum hinges
  • Skin: Lycra fabric (spandex)
  • Pressure sensors measure water flow
  • Performance gap vs real tuna remained unclear
  • Real tuna: accelerate to 40 mph in short bursts
  • Note: 1976 Japanese mechanical carp mentioned but unverified

MIT Evolution Series

VCUV (1998)

  • Vorticity Control Unmanned Underwater Vehicle
  • Evolved from RoboTuna
  • Self-contained autonomous system
  • Field trial speed: 2.8 mph

RoboPike (2000)

  • Experiment in maneuverability without multiple thrusters
  • 32 inches long, 8 pounds
  • Modeled on chain pickerel freshwater fish
  • Challenges: sensor/battery integration, waterproofing electronics

BCF Robot Categories

Anguilliform (Sea Snake-like)

  • Body sashays to generate thrust
  • Example: Envirobot (Swiss EPFL)
    • 4 feet long
    • 6 active modules + flexible tail
    • Quiet monitoring for nature observation

Subcarangiform and Carangiform

  • Body movement in final third only
  • Example: iSplash-2 (University of Essex)
    • Fastest tail-beating fish robot: 8.3 mph
    • Comparison: Michael Phelps swims ~6 mph
    • Limitation: cannot turn

Thunniform

  • Stable body, large crescent-shaped tail only
  • Includes RoboTuna and descendants
  • Examples: tuna, sharks

Ostraciiform

  • Rigid body, constant tail fin flapping
  • Least efficient BCF style but very stable
  • Few moving parts, easier waterproofing
  • Example: Dorybot (Polytechnic University of Marche, Italy)

MPF Robot Categories

Rajiform (Ray-like)

  • Two big side fins flapping
  • Italian cow-nose ray robot
  • Flexible silicone fins with motor
  • Tail serves as rudder
  • Speed: 0.4 m/s (<1 mph)
  • Trade-off: stability/turning vs speed

Amiiform and Gymnotiform

  • Rigid body with long fin wave along top/bottom
  • Like knife fish or electric eel
  • Excellent 3D maneuverability
  • Can go backwards/forwards without direction change
  • Limited development: one prototype in Changsha, China

Labriform

  • Use pectoral fins for maneuvering (like coral wrasses)
  • Switch to tail fins for speed bursts
  • Difficult to build due to low endurance and stability challenges

Soft Body Technology

Shape Memory Alloy (SMA)

  • Most common soft actuator material
  • Metals that remember shape, activated by heat
  • Water environment ideal for cooling, increases motion frequency
  • Various forms: wires, springs

Other Soft Materials

  • Electricity-based soft actuators
  • Chinese snailfish robot example:
    • Control unit/battery/voltage amplifier embedded in polymer
    • Dielectric elastomers as artificial muscles
    • Operated untethered at Mariana Trench bottom
    • Functioned 45 minutes under crushing pressure

Technical Challenges

Computational Power Limitations

  • Ocean environmental noise crowds sensors
  • Complex underwater currents and dynamic obstacles
  • Limited power and space in robot bodies
  • Brain must: observe environment, respond to threats, achieve goals
  • Scientists taking inspiration from real fish for solutions
  • Taipei Tech professor likely working on pathfinding vs swimming mechanics

Real-World Applications

Maritime Monitoring

  • Cheap, quiet, low-power deployment
  • Aquaculture management
  • Trial: 4 swimming fish robots + floating buoy server
  • Data transmission to home base

Industrial Applications

  • Inspection and maintenance
  • Offshore energy industry value
  • Cable communication industry applications

Military Applications

  • Surveillance operations
  • Mine sweeping capabilities
  • DARPA funded MIT's RoboPike project

Conclusion

  • Fish robots represent ideal robotic swimming body design
  • Nature's 500 million years of evolution provides proven designs
  • Performance advantages over traditional propeller systems

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome