Position Sensors for Robotics

By Christian Fell, CTO, POSITAL-FRABA Inc.

Position Sensors for Robotics

Sensor Feedback and the Rise of Collaborative Robots

By Christian Fell, CTO, POSITAL-FRABA

Sensor Feedback and the Rise of...

Thermal Imaging Miniaturization Creating Opportunities in IoT

By Mike Walters, VP, FLIR

Thermal Imaging Miniaturization...

Machine Learning Now at the Forefront of Embedded Applications

By Mark Papermaster, CTO, and SVP Technology and...

Machine Learning Now at the Forefront...

How to Protect Humans from Dangerous Chemical Attacks?

By Enterprise Technology Review | Friday, January 31, 2020

The U.S. Army is developing a tiny wearable sensor to protect against dangerous chemicals.

FREMONT, CA: With the help of Worcester Polytechnic Institute (WPI) mathematician, the U.S. Army is developing a tiny thumbnail-sized chemical sensor to be worn in outer garments to identify dangerous chemicals quickly.

The sensor will mimic the human nose that has roughly 400 types of olfactory receptors to recognize at least one trillion different odors. A combination of multiple molecules can be detected with the help of sensors where each sensor will identify the specific odor. One might detect diesel fumes and humidity while the other might detect diesel fumes and a particular chemical agent. The results are then combined to give a more complete and accurate assessment of the chemicals in the environment. Here data science plays a vital role in advancing significantly chemical sensors such that it will be better able to save lives.

Randy Paffenroth, associate professor of mathematical sciences, computer science, and data science, and a principal investigator on the project, improve the speed and accuracy of the device’s signal processing by combining new and old math techniques. The scientist combined an artificial neural network, which is a new type of machine learning that mimics the function of the human brain and the Kalman filter, a classic algorithm developed to get rid of noisy data to detect problems better while tracking airplanes. Traditional neural networks utilized for this purpose have difficulty with signal processing to detect the presence of chemicals in the air accurately and quickly. This is because when a lot of signals come together and are all noisy, they overlap a lot that could cause false positives or false negatives.

By combining the Kalman filter with the neural network created a new algorithm dubbed the Autoencoder Kalman Filter. The classic and modern algorithms jointly simplify the noise, complex data from the sensors to better detect chemicals in the environment.

Paffenroth has taken advantage of the Kalman filter that shows the power of data science. It’s the toolset that combines data with machine learning, computation, and old and new math to solve real-world problems. It’s the best way of solving a whole class of the issues.

WPI graduate student Matthew Weiss, who is working on the research project with Paffenroth, noted that after analyzing two seconds’ worth of data with the Autoencoder Kalman Filter, the sensor already is five times more accurate than with unfiltered data.

See Also :- Top Enterprise Security Solution Companies

Top