Skip to content

robomechanics/robot-dataset

Repository files navigation

robot-dataset

Robot Scaling Literature Review Dataset Data and figures accompanying the paper: Allometric Scaling Laws for Bipedal Robots. This repository contains morphological data (body length, leg length, mass) for different robot systems spanning locomotion types, used to derive empirical scaling laws via power-law regression.


Repository Contents

File Description
robot_scaling_dataset.csv Robot dataset with morphological measurements and reference URLs
fit_summary_powerlaw.txt Power-law fit results (coefficient A, exponent b, R², n) for all relationships
litrev_mass_length.png Figure: Mass and leg scaling plots (static image)
litrev_mass_length.svg Figure: Mass and leg scaling plots (vector, editable)
litrev_mass_length.html Figure: Mass and leg scaling plots (interactive)

Dataset (robot_litrev.csv)

Columns

Column Units Description
robot_name Robot name
reference Primary reference/source URL
type Locomotion type: biped, quadruped, hexapod
year year Year of publication or commercial release
cost_dollars USD Approximate retail/reported cost
total_length_inm m Total body length
leg_length_inm m Leg length
min_mass_inkg kg Minimum reported mass
max_mass_inkg kg Maximum reported mass
average_mass_inkg kg Average or reported mass used in analysis
normal_speed_inmps m/s Normal operational speed
walking_speed_inmps m/s Walking speed
running_speed_inmps m/s Running/maximum speed
DOF Degrees of freedom
payload_inkg kg Rated payload capacity
actuator_mass_inkg kg Total actuator mass
accuator_count Number of actuators
highest_stall_torqueorforce_on_actuators_inNm Nm Maximum stall torque of primary actuator
actuator_name Actuator model/name
acutation_type Actuation technology (e.g., motors, hydraulics)
reported_COT Cost of Transport
battery_life_hours hours Rated battery life
battery_wt_inkg kg Battery mass

Robot Types

  • Full Biped: humanoid robots with full upper body (n = 17)
  • Lower Biped: bipedal locomotion-only platforms (n = 10)
  • Quadruped: four-legged robots (n = 14)
  • Hexapod: six-legged and insect-scale robots (n = 6)

Scaling Laws (fit_summary_powerlaw.txt)

All relationships follow the form y = A · x^b, fit by ordinary least squares in log–log space.

Key results (global fits across all 47 legged robots):

Relationship A b
Mass vs Body Length 23.76 2.34 0.90
Mass vs Leg Length 68.20 2.12 0.87
Leg vs Body Length 1.48 0.86 0.87

Per-type fits are also provided for Full Biped, Lower Biped, Hexapod, and Quadruped categories.


Notes

  • Data were collected and verified from manufacturer specifications, datasheets, and peer-reviewed publications.
  • Some robots were included in the CSV file, but excluded from powerfitting due to missing data

About

Robot Scaling Literature Review Dataset Data and figures accompanying the paper: Allometric Scaling Laws for Bipedal Robots This repository contains morphological data (body length, leg length, mass) for different robot systems spanning locomotion types, used to derive empirical scaling laws via power-law regression.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages