Various initial ideations of the concept. We decided through discussion to change the configuration so we didn't have to shoot the ball into the air. It would require more power and might break things.
This low-profile design seemed like the best solution for kicking the ball out along the ground and having room to hold a bunch of d batteries. It's also easier to get the ball in the hopper with it taking up most of the top of the device.
I built a mechanism to help us understand how much power and how it might look. We needed to do this to show the factory there were multiple was to solve the problem of how to move the ball. They seemed pretty skeptical the device would work with the kind of low power requirement we wanted for safety.
This is a mechanism suggested by the factory once I built a prototype launcher to show them how it would work.
This is the final configuration for the project.
gLike
Serv-O Ball Launcher Toy

The goal of this project was to entertain a dog while the owner was away. It would require the user to train the dog to drop the ball in the mouth of the device, which would launch the ball (after a slight delay) across the room. Designed to work with a standard tennis ball, it could be set to cease operation after 15 minutes so the dog wouldn't succumb to exhaustion. I was the designer on this project. It was intended to run on D batteries to avoid worry over the dog chewing the power supply cord.

Charles Floyd
Product Development Consultant Oakland, CA