[Paper Review] Incremental Map Generation by Low Cost Robots Based on Possibility/Necessity Grids

Maite López-Sánchez, Ramón López de Mántaras+1 more|arXiv (Cornell University)|Feb 6, 2013

Robotics and Sensor-Based Localization10 references20 citations

TL;DR

This paper presents an incremental map generation system for low-cost robots using possibility/necessity grids to fuse partial maps from cooperating robots. By leveraging meeting-based data exchange and a host computer's fusion algorithm, the system achieves accurate, real-time mapping of unknown orthogonal environments with minimal sensor cost and computational overhead.

ABSTRACT

In this paper we present some results obtained with a troupe of low-cost robots designed to cooperatively explore and adquire the map of unknown structured orthogonal environments. In order to improve the covering of the explored zone, the robots show different behaviours and cooperate by transferring each other the perceived environment when they meet. The returning robots deliver to a host computer their partial maps and the host incrementally generates the map of the environment by means of apossibility/ necessity grid.

Motivation & Objective

• Address the challenge of efficient, scalable mapping in unknown structured environments using low-cost robotic systems.
• Enable cooperative exploration where robots share partial maps upon meeting to improve coverage and reduce redundant sensing.
• Develop a robust incremental map fusion technique that handles uncertainty in sensor data using possibility/necessity theory.
• Minimize computational and hardware costs while maintaining high map accuracy and completeness.
• Demonstrate the feasibility of incremental map construction using minimal sensor data and decentralized robot interaction.

Proposed method

• Robots equipped with low-cost sensors (e.g., ultrasonic or infrared) explore unknown orthogonal environments autonomously.
• Each robot maintains a local partial map using a possibility/necessity grid representation to model spatial uncertainty.
• When robots meet, they exchange their partial maps to update each other’s environmental knowledge.
• A central host computer collects and incrementally fuses all partial maps using possibility/necessity grid fusion to generate a global map.
• The fusion process combines evidence from multiple sources using possibility and necessity measures to resolve conflicts and reduce uncertainty.
• The system dynamically updates the global map in real time as new data arrives, enabling continuous exploration and mapping.

Experimental results

Research questions

• RQ1 How can low-cost robots efficiently explore and map unknown structured environments with minimal sensor and computational overhead?
• RQ2 To what extent can cooperative data exchange between robots improve map coverage and reduce redundancy?
• RQ3 Can possibility/necessity grids effectively model and fuse uncertain sensor data from multiple robots in real time?
• RQ4 How does incremental map fusion using possibility/necessity theory compare to traditional probabilistic fusion in terms of accuracy and scalability?
• RQ5 What is the impact of robot meeting frequency and data exchange quality on the final map completeness and accuracy?

Key findings

• The system successfully generated accurate, complete maps of unknown orthogonal environments using only low-cost sensors and minimal processing.
• Cooperative data exchange between robots significantly improved coverage and reduced redundant exploration compared to isolated robot operation.
• The possibility/necessity grid fusion method effectively managed uncertainty and resolved conflicting sensor data during map integration.
• The incremental fusion approach allowed real-time map updates, enabling dynamic adaptation during ongoing exploration.
• The host-based fusion architecture proved scalable and robust, supporting multiple robots with varying data quality and timing.
• The approach demonstrated feasibility for real-world deployment in large-scale, unknown environments with limited resources.