HermesTrack

A seamless array of RFID sensors, software, and analysis libraries for mapping real-time location data.

Comment

Inspiration

Our fellow team member Ben Lahm recently broke his foot and was admitted in a hospital where he soon realized how lucky he was to have care-providers that enforced sanitary practices such as washing their hands every time they enter and leave the room, etc. Many hospital systems and remote camps do not enforce such miniscule yet essential steps in their system, leading to high infection rates. We wanted to create a way to track such activities for providing real-time feedback correcting bad practices. Realizing the potential for expanding such tracking capabilities and surprisingly finding no easy way to integrate such temporary, cheap systems without hiring professionals, we decided to create our own array of sensors and accessible software libraries that can easily collect and provide data required to implement such corrective feedback and other research analytics.

What it does

HermesTrack utilizes a network of RFID Reader nodes and a central processing Hub node to track the location of users (RFID chips) throughout an enclosed system, like a hospital or office building. This network of sensors allowed for the creation of a cloud database to store real-time data where the users are within the system and therefore activities they are most likely performing (i.e washing hands, talking to patients, using a computer). By utilizing a wrapper-library, users are able to access the data and create applications of their own with this IoT style system. Our example outputs activities of an user of interest with the intention of providing feedback on Mal/bad practices.

How we built it

The hardware employs a network of Raspberry pis for communcation and Arduinos to control the RFID readers. These sensors are all static, and remain in the same location to give us a reference point to locate the mobile RFID chip users. When a user trips one of the sensors, information on user, location, and time is transfered back to a central Hub that processes the incoming data and determines updated locations for all RFID chips in the system.

The software system is meant to act as a processing system and resource for applications to use RFID tracking technology. After the data is processed and the locations of all of the RFID chips are estimated, this data is uploaded to a database for persistent storage. A time stamp associated with each system snapshot can be used to track people or objects as they move. There is then an API that lets users access the data for data analysis, decision analytics, and general purpose tracking.

Challenges we ran into

  • Integrating sensors directly with the Raspberry Pi for input and feedback output - ended up using arduinos for actual data collection
  • Building a more appealing GUI
  • Integrating real-time data from cloud into data-processing software/ Matlab
  • Raspberri internet connection with Illinois-Guest

Accomplishments that we're proud of

  • We were able to successfully collect unique identifier data from RFID sensors and integrate arduino data into Raspberry Pi even though the libraries for the sensors are quite complex, hard to understand and edit
  • We were able to process the data in a central Pi to update cloud whenever there are changes - this is our first time creating a database and updating it real-time in cloud
  • We had to integrated information from multiple different languages that the various team members were comfortable writing in and coordinated data outputs to seamlessly integrate the various parts of the hardware-software system
  • We build our first ever GUI to visually track movement within a hospital room setting

What we learned

  • Remember to bring wifi-shields
  • If something does not work for 4 hours, move on and try different strategies
  • Ask for help! We did not utilize our mentors as much as we should have.

What's next for HermesTrack

The few next steps for HermesTrack.. * Upgrade our sensors to increase the range and usefulness of our sensors * Implement in a larger setting to determine the use of an array of RFID Reader nodes * Continuing the addition of library functions to grab important data from the cloud * Creating a drag-and-drop GUI for demoing the software * Finding new applications of our RFID Reader Tracking Technology

Built With

Share this project:

Updates