[Paper Review] CAT(0) geometry, robots, and society

Federico Ardila|arXiv (Cornell University)|Dec 18, 2019

Digital Image Processing Techniques14 references4 citations

TL;DR

This paper establishes that the configuration space of a 2D pinned robotic arm in a rectangular tunnel is a CAT(0) cubical complex, enabling optimal motion planning via geometric and combinatorial algorithms. The key contribution is a provably optimal algorithm for moving the robot between configurations, with implementation and educational outreach in Colombia highlighting ethical implications of robotic technology.

ABSTRACT

How do we move a robot efficiently from one position to another? To answer this question, we need to understand its configuration space, a 'map' where we can find every possible position of the robot. Unfortunately, these maps are very large, they live in high dimensions, and they are very difficult to visualize. Fortunately, for some discrete robots they are CAT(0) cubical complexes, a family of spaces with favorable properties. In this case, using ideas from combinatorics and geometric group theory, we can construct a 'remote control' to navigate these complicated maps, and move the robots optimally. Along the way, we face larger ethical questions that we cannot ignore.

Motivation & Objective

• To establish that the configuration space of a discrete 2D robotic arm in a rectangular tunnel is a CAT(0) cubical complex.
• To develop an optimal algorithm for moving the robotic arm between any two configurations using geometric and combinatorial techniques.
• To connect mathematical theory with real-world ethical implications, especially in light of the 2016 Dallas police robot incident.
• To broaden access to robotics education by implementing the model in outreach workshops with Colombian students.
• To provoke reflection on the ethical responsibilities of mathematicians and scientists in shaping technological applications.

Proposed method

• Model the robotic arm as a sequence of unit-length links constrained to move in a 2D rectangular tunnel with fixed base.
• Define two local moves: flipping a corner (interchanging directions of two consecutive links) and rotating the end link by 90°.
• Construct the configuration graph where nodes represent arm positions and edges represent valid local moves.
• Prove that the configuration space forms a CAT(0) cubical complex by showing that all 4-cycles correspond to commuting moves and that the space has non-positive curvature.
• Use the CAT(0) structure to apply algorithms from geometric group theory for optimal pathfinding in ℓ₁ (cost) and ℓ∞ (time) metrics.
• Implement the algorithm in Python, providing distance computation and animated optimal trajectories, with public code and educational outreach via workshops.

Experimental results

Research questions

• RQ1 Can the configuration space of a 2D pinned robotic arm in a rectangular tunnel be shown to be a CAT(0) cubical complex?
• RQ2 What algorithmic techniques can be used to compute the optimal path between two configurations in such a space?
• RQ3 How can the geometric and combinatorial structure of the configuration space be leveraged to ensure optimality in motion planning?
• RQ4 What ethical responsibilities do researchers have when developing mathematical tools with potential military or surveillance applications?
• RQ5 How can mathematical research be made more accessible and inclusive through community-based educational outreach?

Key findings

• The configuration space of a 2D pinned robotic arm in a rectangular tunnel is a CAT(0) cubical complex, as proven in Theorem 2.
• An explicit optimal algorithm exists for moving the robotic arm between any two configurations, with both ℓ₁ (cost) and ℓ∞ (time) distances computable.
• The algorithm was implemented in Python and made publicly available, including animations of optimal trajectories.
• The research inspired educational workshops in Colombia, where students designed and prototyped robotic arms for local needs such as firefighting and landslide rescue.
• The work led to a prototype robot built by student Arlys Asprilla using CAD and mechanical design, demonstrating real-world application potential.
• The authors reflect that mathematical tools are not neutral, and that researchers must confront power, funding, and ethical use in scientific practice.