A Benchmark Dataset for Collaborative SLAM in Service Environments

TL;DR

A new multi-modal collaborative SLAM (C-SLAM) dataset for multiple service robots in various indoor service environments, called C-SLAM-SLAM dataset in Service and E environments, and a detailed tutorial on generating C-SLAM data using the simulator is provided.

Abstract

As service environments have become diverse, they have started to demand complicated tasks that are difficult for a single robot to complete. This change has led to an interest in multiple robots instead of a single robot. C-SLAM, as a fundamental technique for multiple service robots, needs to handle diverse challenges such as homogeneous scenes and dynamic objects to ensure that robots operate smoothly and perform their tasks safely. However, existing C-SLAM datasets do not include the various indoor service environments with the aforementioned challenges. To close this gap, we introduce a new multi-modal C-SLAM dataset for multiple service robots in various indoor service environments, called C-SLAM dataset in Service Environments (CSE). We use the NVIDIA Isaac Sim to generate data in various indoor service environments with the challenges that may occur in real-world service environments. By using simulation, we can provide accurate and precisely time-synchronized sensor data, such as stereo RGB, stereo depth, IMU, and ground truth (GT) poses. We configure three common indoor service environments (Hospital, Office, and Warehouse), each of which includes various dynamic objects that perform motions suitable to each environment. In addition, we drive three robots to mimic the actions of real service robots. Through these factors, we generate a more realistic C-SLAM dataset for multiple service robots. We demonstrate our dataset by evaluating diverse state-of-the-art single-robot SLAM and multi-robot SLAM methods. Our dataset is available at https://github.com/vision3d-lab/CSE_Dataset.

Figures (6)

• Figure 1: Illustration of the CSE dataset in Office environments. The CSE dataset is obtained from realistic service environments, including multiple dynamic objects, indicated by a red circle. Our environments contain diverse characteristics, and each box shows the features as seen from the robot camera view. (a) Robot-to-robot interaction ( Follow). The blue circle is the robot driving in front of it. (b) Occlusion from dynamic objects. (c) Complex illumination due to reflective material. (d) Dynamic objects performing motions suitable for the environment.
• Figure 2: Robot configuration and sensor data example. (a) The NVIDIA Carter, our robot platform. (b) Examples of acquired sensor data (stereo RGB, stereo depth, GT poses and IMU).
• Figure 3: Example of service environments in the proposed CSE dataset. Each row shows the service environments we built (Hospital, Warehouse, and Office in order) from several viewpoints. Odd columns represent static environments, while even columns represent dynamic environments. In particular, we can observe dynamic objects having suitable actions and clothes for each environment.
• Figure 4: Challenging cases in the CSE dataset. (a) Occlusions from dynamic objects. (b) Place recognition failure due to similar structure at different location. (c) Invalid feature matching due to a dynamic object at different times. (b) and (c) are cases where SLAM failed.
• Figure 5: Scenarios in the proposed CSE dataset. We visualize scenarios for each dynamic environment on its 2D occupancy map with the same scale. Note that scenarios in this illustration only show dynamic environments.