Using Sensors and Predictive Analytics to Monitor Horse Behavior
- On 10. Mai 2022
Guest Article | Author: Olga Lysak, AI Business Development Manager & Representative Hamburg | Lemberg Solutions
HorseAnalytics is an innovative German startup that uses sensors and predictive analytics to monitor horses‘ behavior and deliver unique equestrian training experiences. This smart application aims to improve horses‘ quality of life by controlling their wellbeing 24/7 with sensors and integrated data analysis.
© Lemberg Solutions
How it all started
The startup founders decided to build a solution enabling round-the-clock collection of data about a horse’s exercise, sleep, and other activities, which would be automatically transformed into meaningful input for equestrians. HorseAnalytics delegated this project to Lemberg Solutions, as we had a proven track record in successful implementation of machine learning projects.
To reach the desired result, our team built a detailed solution architecture with a machine learning algorithm for horse activity detection at its core. Lemberg Solutions was responsible for the all-round technical development, including business analysis, architecture design, mobile app and cloud development, data collection and processing, information model development, and neural network training.
Within a year, the system was ready for public use and made a good impression on customers — we received tons of positive feedback from end-users of the HorseAnalytics app.
© Lemberg Solutions
Zooming in on the development process
We decided to adopt the Lean methodology for the development process and started with building an emulator app instead of developing hardware in the first place. This app would run on regular smartphones located on the animal to gather information about its movement.
At this point, we started building a neural network using the collected raw data. It was crucial to detect noise and discard invalid input, as well as determine the most appropriate format for storing information.
© Lemberg Solutions
We used Google Firebase to collect data from the emulator application. To make sure the assembled input represented relevant movements, we built a voice assistant informing the horse rider about stunts the horse should perform at a given moment.
Simultaneously, we were developing a UI/UX design for the mobile app. We also used MS Azure IoT to set up a cloud environment to facilitate our neural network and improve data storage.
The initial version of the released application could detect four horse activities:
- Standing,
- Walking,
- Trotting,
- Galloping.
We extended this feature offering and upgraded our machine learning algorithm in the following app releases. Besides, our team enjoyed the unique QA process on this project, as it required direct monitoring of the app performance out in the field with horses.
© Lemberg Solutions
We’re happy that Enri Chantal Strobel, HorseAnalytics Founder and CEO, was pleased with our cooperation. „With Lemberg Solutions, I really like the supportiveness. They don’t just develop what you tell them to, but rather try to understand the intention you have for a feature. We have had such a good experience and can definitely recommend Lemberg Solutions,“ said Enri.
Weiterführende Informationen
Kontakt
OLGA LYSAK
AI Business Development Manager & Representative in Hamburg | Lemberg Solutions
Ihre News in unserem Blog
Wir sind immer auf der Suche nach interessanten Inhalten aus den Bereichen (I)IoT, KI, Software, Hardware und Konnektivität sowie den IoT-Anwenderbranchen. Gern veröffentlichen wir auch Ihren Gastbeitrag in unserem Blog.
Luisa Göhler
Digital MarketingKontaktieren Sie mich!
0 comments on Using Sensors and Predictive Analytics to Monitor Horse Behavior