Oct 11 2022Reviewed by Emily Henderson, B.Sc.
A team of researchers supported by the Swiss National Science Foundation (SNSF) and working in the field of navigation has formulated techniques that enable several economical inertial sensors functioning in combination to substitute a single costly sensor.
Inertial sensors are used almost everywhere. Image Credit: © CC-BY-NC-ND: Philipp Foehn, University of Zurich.
Inertial sensors like gyroscopes and accelerometers are used almost in all fields, from smartwatches to vacuum cleaners, drones, submarines, spacecraft, and even game controllers. These sensors indicate the speed, position, or direction of an object. Their downside is the absence of precision, especially in low-cost models used in several devices.
However, owing to support from the SNSF, there may be a way to solve this issue. In an article published in IEEE Transactions on Signal Processing, researchers from the University of Geneva state that networking many economical sensors is a feasible option than more robust sensors.
Using All the Information
By integrating the measurements of several economic, individual sensors, the team successfully obtained a very accurate navigational measurement.
It is as if we had created a virtual sensor, and it is particularly efficient because it uses all the information provided by the individual sensors.
Yuming Zhang, PhD Candidate, Statistics and Study Lead Author, University of Geneva
Offering the same performance as actual sensors, these virtual sensors also cost less and can be more easily set up. They could therefore be employed in several consumer devices without escalating their cost. The concept of merging information from various sensors is not new, but until now, technical limitations have restricted its application.
Sensor measurement errors are very complex to handle individually, and even more so when several sensors are combined.
Yuming Zhang, PhD Candidate, Statistics and Study Lead Author, University of Geneva
The researchers solved the issue using a new signal decomposition method. This method renders it feasible to understand and tackle errors that impact measurements and to process them using a unique statistical technique.
From Aerial Mapping to Finance
According to the Geneva team, the prospective applications of the new method extensively range from aerial mapping with the help of drones to autonomous vehicles. Furthermore, optimally integrating various sensor technologies could help advance a new generation of global positioning devices.
The results could also be used in very distant fields, such as finance. Investment decisions frequently are in the form of a mixture of securities and financial products intended to accomplish a specific goal.
In this field, the method we are proposing could also be used to create an optimal investment combination that minimizes portfolio volatility.
Yuming Zhang, PhD Candidate, Statistics and Study Lead Author, University of Geneva
Journal Reference:
Zhang, Y., et al. (2022) Scale-wise Variance Minimization for Optimal Virtual Signals: An Approach for Redundant Gyroscopes. IEEE Transactions on Signal Processing. doi.org/10.1109/TSP.2022.3208733.
Source: https://www.snf.ch/en