Skip Navigation
Search
Back to Search Innovations
Real-Time Multi-Tag Tracking Process Via a Single Magnetometer Sensor
Magnetic tags worn by drivers to track and classify their head movements through the use of analytical and approximation sensing models

 

Source: Jannis Lucas, https://unsplash.com/photos/3_Pm95bUwLg, Unsplash License

 

 

Background

Every day in the U.S., there are 9 deaths and 1,153 injuries caused by drivers who are distracted behind the wheel. Data from the National Highway Traffic Safety Administration shows that 94% of all car accidents are a result of a driver’s failure to follow safe driving guidelines due to manual and visual distractions, and drowsy driving, Thus, many car collisions can be prevented if there was a way to detect these unsafe driving behaviors in real‑time. Current technologies that combat the inadequacy of drivers behind the wheel include the detection of abnormal vehicle motions, motion detectors, and methods to alert the driver when a dangerous situation arises, There have also been visibility based systems created to detect visual distractions and drowsy driving. Finally, there has also been wearable technologies developed to counter unsafe drivers that measure erratic steering wheel movements and manual distractions. All these conventional technologies have their drawbacks, however. Motion detectors do not allow a sufficient amount of time for the driver to respond to the danger that was detected. Visibility technologies, such as car cameras, are restrained by visibility requirements and deteriorate under bad weather conditions. They also don’t account for manual distractions and aggressive driving. The wearable technologies that have been developed are limited to only one hand. Therefore, a new, novel, reliable, and low‑cost technology is needed to monitor various driver activities

Technology

This technology focuses on reliably monitoring both of a driver’s hands and head motions in real‑time. This method can be applied regardless of the vehicle type and environmental settings and it detects a very wide range of unsafe driving activities from manual and visual distractions to drowsy or aggressive driving. The technology uses battery‑free off‑the‑shelf magnets to facilitate fine grained tracking of the driver’s movements and the design revolves around wearable user‑friendly magnetic accessories for the driver. These wearable magnets pose as active tags that provide signals that convey detailed information regarding the driver’s posture and motion. In association with the wearable magnets, a novel sensing algorithm is used to track the magnetic wearables via a single magnetometer that is on the driver’s smartwatch, Knowing that each magnetic wearable has its own unique patterns of motion and constraints, the algorithm is able to differentiate between them and even track their concurrent motions. Then, based on the results of the tracking, the algorithm is further developed due to its deep learning aspects.

Advantages

  - Independent of weather conditions and visibility constraints - Wearable magnets can be worn on any part of the body the user wants to, and they also measure everything from abnormal hand movements to manual distractions and even aggressive steering - It is reliable, robust, novel, and cost‑effective - 87% precision and 90% recall rate on unsafe driving activities detection in extensive road tests with 500+ instances of different driving activities and 500+ minutes driving time from 10 subjects  

Application

This technology can be applied to everyday drivers as a safety measure based on the driver’s choice. Mandatory use can also be established for drivers that have a history of unsafe and aggressive driving. Type of vehicle, environmental conditions, and driver details do no lower the effectiveness of this technology.

Inventors

Shan Lin, Assistant Professor, Electrical Engineering

Licensing Potential

Development partner - Commercial partner - Licensing

Licensing Status

Available for Licensing.

Licensing Contact

James Martino, Licensing Specialist, Intellectual Property Partners, james.martino@stonybrook.edu,

Patent Status

- Provisional patent - Patent application submitted  

Tech Id

050-9078