Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The Robotability Score: Enabling Harmonious Robot Navigation on Urban Streets

Matt Franchi, Maria Teresa Parreira, Fanjun Bu, Wendy Ju

TL;DR

The paper introduces the Robotability Score ($R$), a metric that quantifies urban environments' suitability for autonomous wheeled sidewalk navigation by aggregating normalized features with polarity-aware weights. Weights are derived from expert pairwise comparisons via an eigenvector method, enabling $R$ to be computed on a sidewalk graph and compared across neighborhoods. NYC-based validation includes feature extraction from OpenData and dashcam analytics, plus four TrashBot deployments that qualitatively corroborate the score’s predictive utility. The framework is designed to be extensible and open-source, aiming to guide harmonious robot deployment without supplanting human-centered urban planning.

Abstract

This paper introduces the Robotability Score ($R$), a novel metric that quantifies the suitability of urban environments for autonomous robot navigation. Through expert interviews and surveys, we identify and weigh key features contributing to R for wheeled robots on urban streets. Our findings reveal that pedestrian density, crowd dynamics and pedestrian flow are the most critical factors, collectively accounting for 28% of the total score. Computing robotability across New York City yields significant variation; the area of highest R is 3.0 times more "robotable" than the area of lowest R. Deployments of a physical robot on high and low robotability areas show the adequacy of the score in anticipating the ease of robot navigation. This new framework for evaluating urban landscapes aims to reduce uncertainty in robot deployment while respecting established mobility patterns and urban planning principles, contributing to the discourse on harmonious human-robot environments.

The Robotability Score: Enabling Harmonious Robot Navigation on Urban Streets

TL;DR

The paper introduces the Robotability Score (), a metric that quantifies urban environments' suitability for autonomous wheeled sidewalk navigation by aggregating normalized features with polarity-aware weights. Weights are derived from expert pairwise comparisons via an eigenvector method, enabling to be computed on a sidewalk graph and compared across neighborhoods. NYC-based validation includes feature extraction from OpenData and dashcam analytics, plus four TrashBot deployments that qualitatively corroborate the score’s predictive utility. The framework is designed to be extensible and open-source, aiming to guide harmonious robot deployment without supplanting human-centered urban planning.

Abstract

This paper introduces the Robotability Score (), a novel metric that quantifies the suitability of urban environments for autonomous robot navigation. Through expert interviews and surveys, we identify and weigh key features contributing to R for wheeled robots on urban streets. Our findings reveal that pedestrian density, crowd dynamics and pedestrian flow are the most critical factors, collectively accounting for 28% of the total score. Computing robotability across New York City yields significant variation; the area of highest R is 3.0 times more "robotable" than the area of lowest R. Deployments of a physical robot on high and low robotability areas show the adequacy of the score in anticipating the ease of robot navigation. This new framework for evaluating urban landscapes aims to reduce uncertainty in robot deployment while respecting established mobility patterns and urban planning principles, contributing to the discourse on harmonious human-robot environments.

Paper Structure

This paper contains 60 sections, 11 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Urban sensing
  4. Scoring the urban environment
  5. Robot deployments in urban streets
  6. Methods
  7. Expert Interviews
  8. Survey for Pairwise Comparison
  9. Derivation of Weights
  10. Feature Set and Pairwise Combinations
  11. Pairwise Rankings and Contingency Matrix
  12. Eigenvalue Decomposition for Weight Calculation
  13. Computing the Robotability Score, $R$
  14. Feature Normalization and Interpretation
  15. Summation with Polarity Control
  16. ...and 45 more sections

Figures (3)

  • Figure 1: Robotability Score distribution in New York City. Blocks that are colored white indicate a lack of sidewalks, or a lack of dashcam data to estimate features. Red dots indicate the locations of our two deployments in areas of lowest $R$, and green dots indicate the locations of our two deployments in areas of highest $R$. The weights used to generate this distribution are in the "NYC POC" column of \ref{['tab:weights']}. A larger map, displaying the robotability distribution from the perspective of TrashBot's requirements, can be viewed in \ref{['fig:rs-nyc-trashbot']}.
  • Figure 2: Trashbot deployments in two of the highest and two of the lowest robotability neighborhoods in NYC.
  • Figure 3: Robotability Score distribution in New York City, computed with the set of weights specialized for TrashBot, shown in \ref{['tab:weights']}. Blocks that are colored white indicate a lack of sidewalks, or a lack of dashcam data to estimate foundational features. Red dots indicate the locations of our two deployments in areas of lowest $R$, and green dots indicate the locations of our two deployments in areas of highest $R$.