Introduction

This project earned my group an “A” in undergraduate Senior Design. The premise of Senior Design class is that a group of students must design and build an original product. We arrived at this unusual idea because one group member was a former horse owner and our professor’s contribution was “make it web-enabled.” Thus, we had an idea that was truly original and we set out to build it. A little background information on horse feeders will help to clarify the project.

Current Feeders

Current automated horse feeders use mechanical and/or electrical timing devices to dispense feed at regular intervals. Power is usually supplied through a 12-volt battery, a 110-volt outlet, or even solar power. Generally, there are two mechanisms: one is a timer and the other a measurer. When the timing mechanism reaches the preset times, a motorized auger turns allowing grain to be dispensed. Timing mechanisms are somewhat simplistic and provide limited programming capability.

All feeders have override buttons to dispense grain manually. Few, if any, provide the ability to remotely control the parameters of a particular feeder. The user is limited to programming the device from the feeder itself. Every feeder that uses the current programming system must be individually programmed.

Current systems range in price with a low-end model selling for $285 and a high-end model for $865. No web-enabled models were found.

Our Feeder

The problems with current feeders centered on usability, scalability, and accessibility. After recognizing the problems/limitations of current feeders, our task was to find and implement a solution set. Our project addressed each of these concerns:

• Usability — we improved the range of usability and programability with a Z-World microcontroller. Using a multi-line LCD display, keypad, and the microcontroller, we expanded upon the usability constraints of other feeders. Users could log into the feeder using a password, program feed times, frequency, and feed quantities. Thus, the feeder could function independently from an internet connection. The microcontroller was coded in Dynamic C.
• Scalabilty — the server acts as the heart of the system. Users can register and control every feeder from one online location. In theory, the system can support an unlimited number of horse feeders. The only limitations are server speed and storage capacity. We also added a camera to the system. Thus, the user could see the horse (if it was near the feeder), set feed times and amounts, and view a log file of all feeder events. The server was running Red Hat Linux and a MySQL database. The microcontroller interface was coded in C and C++. The web interface was coded entirely in PHP.
• Accessibility — as mentioned earlier, the system allows the user to program the feeder locally (at the feeder itself) or remotely (from the web site). Although feed times will vary infrequently, the concept proves that such a system increases the horse owner’s ability to monitor what is happening.

The following diagram illustrates the basic operation of our system:

Marketing

From a marketing standpoint the web-enabled, automated feeder could be sold as a service. Consider a potential customer looking for a way to fill his horse stalls. He buys several feeders and subscribes to the online service. He then distributes feeders to paying clients who want to care for their own horses. This fee provides the clients with a horse feeder and all the remote accessibility features.

Conclusion

This project was both fun and educational. It also proves that such a system is viable and practical for a segment of the horse-owning population. Having an excellent team also helped to make this an enjoyable experience. Our test subject, Reno, also thought highly of our project.

Appendix

Many of these images were created for documentation purposes. The help section of Reno Server provided these screen captures with a thorough explanation of each. The feeder itself was also well documented.

• Feeder Hardware: The feeder hardware was donated to our group from a local horse owner and engineer located in Sanford, FL. We were very grateful because this gave us the necessary mechanical components and allowed us to focus on the computer engineering aspects of the project.
• Reno: Our sponsor allowed us to visit his horse, which immediately became our group mascot. It’s always good to meet the end user.
• Login: This page welcomes every user to the Reno Server.
• Registration: This step is required to use the service.
• User Page: Once the user has added a feeder to the system this page will display summary information. The system allows n number of feeders per user, which is great for a stable.
• Feed Times: Times are displayed and can be updated, deleted, or added to each feeder. Quantity can be varied for each feeding time.
• Feed History: All events are logged in the database.
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 

This entry was posted by Matthew Crumley on Saturday, August 4th, 2007 at 8:53 AM. Matthew is currently a software developer who works on simulation devices and their user interfaces. In his spare time he reads the news, dabbles with code, is active in his homeowner association, and enjoys life with his wife and pet terrier. You can leave Matthew a comment or trackback from your own site. Follow comments to this entry via RSS. Explore other posts marked with these tags: college, project, web design.