Analysis
During our visit to the Children’s Museum, we discovered some limitations of our project. Firstly, the camera didn’t track the ball well due to the change of light. It requires calibration of the color threshold every time when the lighting condition is changed. Secondly, the pulley system and the motor were not functioning well, making the goalie not able to react quickly to the ball. Third, the programmed goalie motion was not responsive or too simple.
Despite the limitations, children and adults actually enjoyed the project in some ways. Mostly adults and elder children were able to notice the purpose of the project. For younger children, they usually didn’t know what the machine does initially. Once they realized it was a soccer or hockey game, they would try to shoot the ball. They also tried to “kick” in different directions or used the side rail to reflect the ball and fake the goalie.
However, one aspect of the interactions surprised me. Children actually “kick” the ball very gently. Initially, we thought they would try as fast as possible, even throw the ball to the goalie, however, they threw very gently. Some younger kids were not able to make the ball to reach the goal. Another aspect was that many children felt the game was too easy. They actually wanted to have a challenging game. Instead of scoring for most of the time, they would enjoy more if the goalie was able to block more. Also, some children tried to “cheat” in the game. They placed the ball right in front of the goal and poked it, making the goalie not able to react to the shot.
To improve the interaction with the project, we could firstly aesthetically improve the goalie and play field. Currently, the goalie is representative of what it does, but hardly looks like an actual goalie. Additionally, there is no indication of a ‘D’ or where to place the ball to start (hence the children ‘cheating’ the game). The goalie could be made to look more like a human than a simple object, and arrows could prompt the direction of motion.
When the ball cleared the goalie, there was no confirmation that the player had actually scored without observing the balls placement in the goal. To improve this, we could install LEDs that shine when the user scores to remove any ambiguity. These LEDs will be signaled when a sensor detects the ball’s presence inside the goal.
Lastly, another feature that could be implemented is a simple obstacle the player must navigate the ball through first to then get to the goal, for example a horizontal beam placed across the playing field. This will not only reduce the incoming ball speed giving the goalie more time to react but also increase the challenge and interactivity of the game.
Revision Plan
In order to solve the camera tracking problem, we will implement an auto-tune function for the tracking. So for each time it starts, it searches for the ball and extract the color and uses the new color as the threshold for this run. Secondly, we might need to try different pulleys or redesign the pulley system. Also, we are going to try batteries since the power outlet is not able to supply higher power to the motor. Third, we are going to test using the motor encoder to track the goalie’s position instead of the two switches and reduce the camera tracking latency to make the goalie more responsive.
We think the fundamental experience was quite successful during the visit, so we are going to keep it as what it was, a table soccer game.
Demo Video
Leave a Reply
You must be logged in to post a comment.