A Cooperative Positioning Framework for Robot and Smart Phone Based on Visible Light Communication

Table of Contents

1. Overview

This paper addresses the challenge of indoor positioning where traditional technologies like GPS fail due to signal blockage. It proposes a cooperative positioning framework leveraging Visible Light Communication (VLC). The system uses LED lights modulated with On-Off Keying (OOK) to transmit identifier (ID) and position data. A smartphone's CMOS camera, utilizing the rolling shutter effect, captures these light signals as stripes, enabling high-speed Optical Camera Communication (OCC). By decoding these stripes, the device retrieves a Unique Identifier (UID) linked to a pre-mapped physical location, thereby determining its own position. The framework is designed for scenarios requiring human-robot collaboration, such as warehouses and commercial services, where real-time, shared location awareness is critical.

2. Innovation

The core innovation lies in the design of a unified VLC-based system for cooperative positioning between smartphones and robots. Key contributions include:

1. Multi-Scheme VLP Design: The system incorporates several Visible Light Positioning (VLP) schemes to handle different smartphone tilt postures and varying lighting conditions, enhancing practical robustness.
2. Integrated Cooperative Framework: It establishes a real-time platform where both smartphone and robot locations are acquired and shared on the smartphone interface, enabling mutual awareness.
3. Experimental Validation: The study focuses on and experimentally verifies key performance metrics: ID identification accuracy, positioning accuracy, and real-time capability.

3. Description of Demonstration

The demonstration system is bifurcated into transmitter and receiver components.

3.1 System Architecture

The architecture consists of LED transmitters, controlled by a Microcontroller Unit (MCU), broadcasting modulated position data. The receivers are smartphones (for human tracking) and robots equipped with cameras. The smartphone acts as a central hub, processing VLC data from LEDs for self-localization and receiving robot position data (potentially via other means like WiFi/BLE) to display a unified, cooperative map.

3.2 Experimental Setup

As indicated in the text (Fig. 1), the setup involves four LED transmitters mounted on flat plates. A scalable control circuit unit manages the LEDs. The environment is designed to simulate a typical indoor space where both a robot and a human with a smartphone operate.

4. Technical Details & Mathematical Formulation

The core technology hinges on OOK modulation and the rolling shutter effect. The LED's on/off state, modulated at a high frequency, is captured by a CMOS sensor not as a uniform bright/dark image but as alternating dark and bright bands (stripes) across the image. The pattern of these stripes encodes digital data (the UID).

Position Estimation: Once the UID is decoded, a lookup in a pre-established database provides the LED's world coordinates $(X_{LED}, Y_{LED}, Z_{LED})$. Using camera geometry (pinhole model) and the detected pixel coordinates $(u, v)$ of the LED's image, the device's position relative to the LED can be estimated. For a simplified 2D case with known LED height $H$, the distance $d$ from the camera to the LED's vertical projection can be approximated if the camera's tilt angle $\theta$ and focal length $f$ are known or calibrated:

$ d \approx \frac{H}{\tan(\theta + \arctan(\frac{v - v_0}{f}))} $

where $(u_0, v_0)$ is the principal point. Multiple LED sightings enable triangulation for more accurate 2D/3D positioning.

5. Experimental Results & Chart Description

The paper states that the framework's feasibility, high accuracy, and real-time performance were demonstrated based on the experimental system. While specific numerical results are not detailed in the provided excerpt, it references achieving high accuracy (e.g., 2.5 cm in related robot-only work [2,3]).

Implied Charts/Figures:

• Fig. 1: Overall Experimental Environment and Result: Likely shows the physical setup with four LED panels, a robot, and a person with a smartphone. A schematic or screenshot of the smartphone display showing real-time positions of both entities on a map would be the key "result."
• Accuracy Evaluation Charts: Typical plots would include Cumulative Distribution Function (CDF) of positioning error for both static and dynamic tests, comparing the proposed method against a baseline.
• Real-time Performance Metrics: A graph showing latency (time from image capture to position display) under different conditions.

6. Analysis Framework: Example Case

Scenario: Warehouse Order Picking with Human-Robot Team.
Step 1 (Mapping): LEDs with unique UIDs are installed at known locations on the warehouse ceiling. A map database links each UID to its $(X, Y, Z)$ coordinates.
Step 2 (Robot Localization): The robot's upward-facing camera captures LED stripes, decodes UIDs, and computes its precise location using geometric algorithms. It navigates to inventory bins.
Step 3 (Human Worker Localization): A picker's smartphone camera (potentially tilted) also captures LED signals. The system's multi-scheme VLP compensates for the tilt, decoding the UID and determining the worker's location.
Step 4 (Cooperation): The robot and smartphone exchange their coordinates via a local network. The smartphone app displays both positions. The robot can navigate to the worker's location to deliver a picked item, or the system can alert the worker if they are too close to the robot's path.
Outcome: Enhanced safety, efficiency, and coordination without relying on weak or congested RF signals.

7. Application Outlook & Future Directions

Near-term Applications:

• Smart Warehouses & Factories: For inventory robots, AGVs, and workers in logistics.
• Healthcare: Tracking mobile medical equipment and staff in hospitals.
• Retail: Customer navigation in large stores and interaction with service robots.
• Museums & Airports: Providing precise indoor navigation for visitors.

Future Research Directions:

1. Integration with SLAM: Deep fusion of VLC-based absolute positioning with robot's SLAM (as hinted in [2,3]) for robust, drift-free navigation in dynamic environments.
2. AI-Enhanced Signal Processing: Using deep learning to decode VLC signals under extreme conditions (motion blur, partial occlusion, interference from other light sources).
3. Standardization & Interoperability: Developing common protocols for VLC positioning signals to enable wide-scale deployment, akin to efforts by the IEEE 802.15.7r1 task group.
4. Energy-Efficient Designs: Optimizing the smartphone-side processing algorithms to minimize battery drain from continuous camera use.
5. Heterogeneous Sensor Fusion: Combining VLC with UWB, WiFi RTT, and inertial sensors for fault-tolerant, high-availability positioning systems.

8. References

1. [1] Author(s). "A positioning method for robots based on the robot operating system." Conference/Journal, Year.
2. [2] Author(s). "A robot positioning method based on a single LED." Conference/Journal, Year.
3. [3] Author(s). "[Related work] combined with SLAM." Conference/Journal, Year.
4. [4] Author(s). "On the cooperative location of robots." Conference/Journal, Year.
5. [5-7] Author(s). "VLP schemes for different lighting/tilt situations." Conference/Journal, Year.
6. IEEE Standard for Local and metropolitan area networks--Part 15.7: Short-Range Optical Wireless Communications. IEEE Std 802.15.7-2018.
7. Gu, Y., Lo, A., & Niemegeers, I. (2009). A survey of indoor positioning systems for wireless personal networks. IEEE Communications Surveys & Tutorials.
8. Zhuang, Y., et al. (2018). A survey of positioning systems using visible LED lights. IEEE Communications Surveys & Tutorials.

9. Original Analysis & Expert Commentary