Software team integral to robotics team’s success

For senior Eugene Che and junior Alex Renda, robotics season means that they have to finish their homework early. From 6 to 9 pm, they are using the coding program Eclipse to write software for the M-SET robot.

“I like to think of robotics programming as human interfacing for the most part — i.e. making it easier for a person to interact with the robot,” Che said. “When developing software for the robot, we consider how to make human input as intuitive as possible. ”

He said that the software team essentially “translates joystick movements into power given to motors” and “writes code to drive the robot during the autonomous period of the game.”

Che said they also consider what would be difficult for a human attempting to accomplish the task the robot is designed to carry out and how they can eliminate those difficulties.
“In addition to that, there is a time restriction of six weeks for when we need to have the code finished,” Che said. “This requires us to manage code on a human level, in that we have train new members, distribute coding assignments effectively, and write code that we don't regret reading in the future.”

Renda said that they continue learning and using machine vision, which allows robots to detect potential targets, this year to program the robot to detect a target and aim at a target so it would know where exactly to throw.

“The most challenging part is by far the machine vision,” Renda said. “Last year, the first year we tried to write machine vision code, our process was essentially trial and error. We tried several methods without much forethought and were ultimately disappointed in the unreliable code we ended up with.”

Renda said that the code the team writes is markedly different every season.

“What we write varies from year to year based on the challenge, but this year we will definitely be focusing on writing code to allow the drivers to more efficiently move the robot [around the field],” Renda said. 

Renda said that this year, the team is planning on learning from the mistakes of previous years.

“We're already starting to brainstorm and research ways to improve both our process and our resulting code, and we hope to create both good code to use for competition this year and an effective design process for creating that code in the years ahead.”