Wi-Fi-Based Indoor Robot Positioning Using Machine Learning
Author Information
Author(s): Abacı Hüseyin, Seçkin Ahmet Çağdaş, Chen Xi, Li Xingxing
Primary Institution: Adnan Menderes University
Hypothesis
Can machine learning improve the accuracy of indoor robot positioning using Wi-Fi signals?
Conclusion
The study demonstrates that a machine learning-based method using Wi-Fi signals can achieve high accuracy in indoor robot positioning.
Supporting Evidence
- The AdaBoost algorithm achieved a mean absolute error of 0.044 m on the x-axis.
- Using only seven Wi-Fi access points, the mean absolute error was still acceptable at 0.811 m.
- The study utilized a dataset collected from 398 measurement points across a four-floor building.
Takeaway
This study shows that robots can find their way indoors using Wi-Fi signals, and they can do it really well with the help of smart computer programs.
Methodology
Data was collected using a Raspberry Pi-based robot that scanned Wi-Fi signals at various points, and machine learning algorithms were applied to predict the robot's position.
Potential Biases
Potential biases may arise from the specific Wi-Fi access points selected and the environmental conditions during data collection.
Limitations
The study may not generalize to all indoor environments due to specific conditions and the reliance on Wi-Fi infrastructure.
Statistical Information
P-Value
0.044
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website