CrowdInside: Automatic Construction of Indoor Floorplans (1209.3794v1)

Abstract: The existence of a worldwide indoor floorplans database can lead to significant growth in location-based applications, especially for indoor environments. In this paper, we present CrowdInside: a crowdsourcing-based system for the automatic construction of buildings floorplans. CrowdInside leverages the smart phones sensors that are ubiquitously available with humans who use a building to automatically and transparently construct accurate motion traces. These accurate traces are generated based on a novel technique for reducing the errors in the inertial motion traces by using the points of interest in the indoor environment, such as elevators and stairs, for error resetting. The collected traces are then processed to detect the overall floorplan shape as well as higher level semantics such as detecting rooms and corridors shapes along with a variety of points of interest in the environment. Implementation of the system in two testbeds, using different Android phones, shows that CrowdInside can detect the points of interest accurately with 0.2% false positive rate and 1.3% false negative rate. In addition, the proposed error resetting technique leads to more than 12 times enhancement in the median distance error compared to the state-of-the-art. Moreover, the detailed floorplan can be accurately estimated with a a relatively small number of traces. This number is amortized over the number of users of the building. We also discuss possible extensions to CrowdInside for inferring even higher level semantics about the discovered floorplans.

• The paper introduces a novel system that uses crowdsourced smartphone sensor data to automatically construct indoor floorplans with over 12x error reduction via an anchor-based technique.
• It employs dedicated modules for data collection, trace generation, and floorplan estimation to accurately distinguish corridors, rooms, and key points of interest.
• The approach achieves impressive accuracy with a 0.2% false positive and 1.3% false negative rate, offering a scalable solution for indoor navigation and location-based services.

Overview of "CrowdInside: Automatic Construction of Indoor Floorplans"

The paper "CrowdInside: Automatic Construction of Indoor Floorplans" authored by Moustafa Alzantot and Moustafa Youssef, presents an innovative approach to the automatic construction of indoor floorplans utilizing data from smartphone sensors through crowdsourcing. This research addresses a significant challenge in indoor location-based applications, which require up-to-date and accurate floorplan data—a functionality that traditional means fail to deliver efficiently.

System Framework and Key Contributions

The "CrowdInside" system capitalizes on smartphone sensors, including accelerometers, gyroscopes, and magnetometers, to generate motion traces of individuals as they navigate within buildings. These traces, which are collected passively and transparently, constitute the core data for estimating building layouts. The paper outlines the three primary modules that comprise the system:

1. Data Collection Module: This module is responsible for collecting sensor data, including inertial sensor measurements and WiFi signal strengths, that are later processed to map indoor spaces.
2. Traces Generation Module: Utilizing a novel anchor-based error resetting technique, this module improves the accuracy of the generated motion paths, delivering a substantial enhancement over existing dead-reckoning methods with an error reduction factor of more than twelve times.
3. Floorplan Estimation Module: This segment of the system discerns between corridors and rooms using classification techniques and further applies clustering and computational geometry methods to furnish a detailed representation of the floorplan.

A notable feature of the CrowdInside approach is its ability to distinctly identify points of interest such as stairs, elevators, and corridors with impressive precision, showcasing a 0.2% false positive rate and 1.3% false negative rate. This accuracy is pivotal in resetting the accumulation of errors in motion traces.

Practical Implications and Theoretical Insights

From a practical perspective, CrowdInside provides a cost-effective and scalable solution for creating indoor floorplans, which can significantly benefit applications like indoor navigation, location-based advertising, and social networking services within buildings. The reliance on widely deployed smartphone sensors means that no additional infrastructure is required, reducing both deployment complexity and cost.

Theoretically, the results highlight the potential of crowdsourced data in constructing complex environmental models, suggesting that similar methodologies could be applied to various domains where sensor data is abundant and detailed environment mapping is required. The paper’s findings could inspire further exploration into integrating semantic information, enhancing energy efficiency, and developing user incentives for data collection.

Future Directions

While the CrowdInside system demonstrates significant progress, it opens avenues for future research. Potential directions include integrating more robust machine learning models to infer higher-level semantic details such as specific room usage or ownership. Additionally, optimizing for energy efficiency remains crucial, especially given the continuous nature of data collection from mobile devices. Exploring decentralized approaches could also enhance user privacy while distributing the computational load.

In conclusion, "CrowdInside" offers a comprehensive framework for harnessing the power of crowdsourcing in mapping indoor environments, marking an important step forward in ubiquitous computing and smart building technologies.