How Important is Speed in ASEE Competitions?

If I’ve told you once I’ve told you one thousand times, speed is the last thing you should be concerned with in an ASEE competition, read on and I’ll prove it.

For starters, it is my professional opinion that speed is the easiest feature to enhance and requires the least mechanical/physical change which is a big part of why it should be saved for last. There are many ways to increase speed that may be an option even after you’ve completely built your robot. Some examples of simple ways to increase speed are shown a little later but first, I’d like to go over the most compelling argument, proof by previous competition results.

Analysis of Past ASEE Competition Scores

Here I will walk through a few of the results over the past five years of competitions. Unfortunately, I do not have the data for every year but the years that I do have will be plenty to show the simple following fact.

You will be lucky to place in the top five if you do not have consistent and repeatable trial runs, the only time that a time bonus will be of any worth is when you are more consistent and reliable than every other team.

ASEE 2011 – Tree Harvester

Notice the far right column, this shows how many of the four trial runs were perfect runs. The top four places are the only teams to achieve this feat. Then we are followed by two teams with three perfect runs and a team with two perfect runs.

ASEE 2012 – Pony Express

Notice, first place happens to be the only team that achieved four perfect runs, followed by the teams with three, two and one perfect run respectively.

ASEE 2014 – Mini Indy

Notice, all top five positions are held by those with four perfect runs. Now, you may be looking at this and saying… ‘hey, team 10 didn’t get any perfect runs and still made it to 6th place’ and you would be right but I’ll explain this.

A perfect run in this game was to achieve 5 laps within a specified time, you received points for each lap and a time bonus if you completed all five in less than the time allotted. This is where it gets good because clearly the teams in positions 7, 8 and 9 all received this time bonus twice but that didn’t matter because it didn’t make up for the fact that team 10 had achieved four laps in every trial which is near perfect. They only achieved four laps because they were too slow to get all 5 in the allotted time… too slow… get it, even though they were too slow, they still beat faster robots.

So, this only serves to prove even further that the time bonus should be the last goal you strive to achieve. The fact that team 10 achieved near perfect runs consistently resulted in a better score than those who had perfects runs that were fast but inconsistent. Each of the three teams with two perfect runs received, on average, 25 time related bonus points each in each trial but still couldn’t place higher than a consistent, close to perfect robot.

ASEE 2016 – Robot Parade

In 2016 we see again, that a team without a perfect run scored better than on with a perfect run. If you’ll notice as well, they’re ‘Test Score’ is 110 points higher than the team in 5th place so they had quite an extensive buffer to hold their place.

So, how did team number 5 get such a lead over team number 16? Team number five placed rings in 10, 11, 11 and 10 boxes each trial respectively and returned to start (receiving a bonus) in every trial. Team number 16 placed rings in 10, 1 12 and 0 boxes each trial respectively and returned to start in two trials. Yet again, this proves that speed and a time bonus means nothing when compared to consistency and repeatability.

Possible Speed Enhancement Methods

Higher RPM Motors

Use motors with higher RPM (assuming the lower torque is not going to be an issue). For a given voltage, speed and torque are inversely proportional, when you increase the RPM while applying the same voltage you will get less torque so keep this in mind. If you are using geared motors, many times you can find a motor with the exact same dimensions that uses a different gear ratio to give you higher RPM with the exact same sized motor.

Bigger Wheels

Use bigger wheels with the same motor (assuming the lower torque is not going to be an issue). Using bigger wheels will decrease the amount of torque you have

Use a combination of the previous two options.

You could find a motor of the same dimensions with a higher RPM and near the same torque but will require a higher voltage. For example, maybe you currently have a 6 volt motor but there is a 9 volt version that provides the same torque with an increase of 30% RPM so you don’t have sacrifice torque for increased speed but you may need to use larger battery.

MCU/Software Enhancements

Sometimes, no matter how hard you try, you can’t get your robot to go faster because it becomes unstable when line following at certain speeds. This could be related to your sampling rate (the speed at which you are reading the line). If you have a slow processor that can only handle reading the line 100 times per second, you might not see a turn until you are three inches into it. An easy way to increase this resolution might be using a faster processor. Another way to increase resolution could be as simple as software enhancements, talk with someone who has intimate knowledge of how to tweak software to take full advantage of the processing power. Anyone in the computer science field with a good majority of their degree finished should be aware of how to do this.

Conclusion

Can you plan for speed enhancements when designing your robot, sure you can, but I stress that you only keep the enhancements in mind. Know that you might need to change the motors or the wheel size or that you might need to use a different MCU board so design in a way that might allow these things to change. Definitely do not focus on these things, there are so many ways to increase speed in a simple way that you should not be concerned with it in the beginning. This is the same case in software development and why we have a name for it called premature optimization. It’s what happens when developers are too overly concerned with what data structure, variable, etc is faster when you don’t even know if it’s going to be a concern, it’s such an easy change later on, don’t worry about it upfront.