Hygiene monitoring and activity tracking system
Barkom Ltd. is a Ukrainian agricultural company that specializes in growing pigs and cattle and selling self-produced food products.
The African swine fever virus is highly resilient and extremely dangerous to domestic pigs. If the virus makes its way onto a pig farm, the only way to contain the outbreak is by culling the entire population.
Barkom takes serious measures to prevent African swine fever and other infectious diseases from entering the farm. Their request for Lemberg was to automate staff hygiene monitoring and control without any major disruptions to the existing biosafety procedures.
Our engineering team analyzed the farm biosafety area, which includes changing rooms and showers and proposed to use beacons and smartwatches to track how proper each employee cleaning procedures are.
The proposed solution would remove any human factors from the hygiene control process. Also, farm managers would be able to view data about staff's hygiene compliance on a dedicated web dashboard and take measures to improve the procedures if necessary.
With an eye to thorough understanding of client needs, we began our work with the research.
We understood that our solution needs not only to track the location of farm staff but also to show whether an employee takes a thorough shower.
After fumbling around with different options, our choice fell on the combination of a beacon device, smartwatch, and remote server for computation. The idea is that the smartwatch will sense the distance to the beacon device, store the accelerometer data, and send everything to the remote server for further computation. Sensing the distance, we will know in what room a person is located. Storing the accelerometer data, we will analyze human movements, and therefore we may arrive at a verdict on how properly employees adhere to biosafety measures.
The very process of development was divided into three stages.
During the first stage, we selected hardware components and built a system demo at the office. We started collecting data on zone entrance and departure and analyzed data accuracy.
Then, at the second stage, we tried to replicate our demo on a farm. It turned out that beacon RF signals interfered with one another in the changing room, and smartwatch RF signals interfered with one another outside the changing room. Therefore, we had to reposition the beacons, use steel sheets as shielding material, and paint walls with conductive ink.
There were a few more special conditions of this solution, like short battery life and employees forgetting to charge their smartwatches. That is why we designed a custom smartwatch charging station. Our vision relied on the duty of employees to leave their smartwatches at the station. While charging, devices transmit the accelerometer data to the server. It prevents constant use of WiFi and, as a result, saves battery life.
Since the hardware was ready, we proceeded to the third stage, where our team of data scientists had to design scenarios of data collection during the actual showering procedure. After that approach was shared with all participants, so everybody had to record their way of showering, the data sets were labeled correctly and ready for further modeling by Lemberg data scientists. The resulting motion analysis model was reflected on a dashboard - a platform for at-a-glance viewing of showering analytics.
Published by Lemberg Solutions on Sep 7, 2020