I was having a discussion today with someone in the robotics industry today that brought to the forefront a question that had been in the back of my mind for a while. The question is “Are we really building robots?” The issue being that our creations are primarily teleoperated.
He suggested that by emphasizing the mechanical and electrical aspects of our creations to the detriment of software that FIRST and associated vendors are being harmful and preying on the ignorance of the average high school student to what a “robot really is”.
I am interested in hearing the thoughts of others on this topic. I’ll wait a bit before I post mine here.
Yes we are building robots but in a simple way. We are using sensors and other things to build objects that move on their own. We have a autonomous period too. Could we do better? I think so otherwise you could call it an expensive and very"smart" (can’t think of another word) remote control car. My dad installs AS/RS machines and even though mostly autonomous they have user control and they way they go about it isn’t much different than the way we do, we just have different operations. If you don’t want to admit it’s a full robot than maybe a “dumb” robot?
This has been brought up before, a few years ago. There’s a couple of old threads on the topic. (I’m just too lazy to go look them up right now.)
Now, I’m going to counter that question with a simple question: If an industrial robot (say, for example, an automotive welder) is taught what it is supposed to do by being guided through its motions and then told, “OK, go do this and tell us if you break down” is not a robot when it is guided through its operation (because that involves direction from an operator), then should we say that whoever is marketing it as a robot is lying?
Or how about the Mars rovers? Do those count? After all, they are guided by human controllers here on earth. They just have a lot of autonomy over what they do–but they’re still controlled.
I could go on with the list–PackBot and Talon come to mind, as do some flying unmanned “robots”–but if you’re going to call our creations not robots, then you need to be prepared to call out all of the above as not robots as well. Just because something is remotely operated does not mean that it is not a robot.
Either that, or everybody is being fooled, not just high school students.
It may also be that the term “robot” has a wider meaning these days, probably because there isn’t a nice short term for “smallish remotely-controlled device that moves around and does stuff”.
A common definition of “robot” is a machine that that senses the world around it and uses this information to interact with the world in a way that it appears intelligent to an observer. So by this definition, most FIRST “robots” are not robots as w/o a constant stream of human input, they won’t do much. I agree with the term “glorified RC cars.”
Ok, so the general consensus seems to be trending towards most teams (of course there are always exceptions) building glorified RC cars. So the next question is, “Is that a bad thing?” Are we doing a disservice to the industry and to the students by promoting this as a robot competition?
What IS a robot? Are the DARPA Vehicles a robot? Is a military drone a robot? The only thing that I can think of that I would classify as a robot 100% are the ABB / Fanuc / etc robots we have in our plant. They are built using the same basic components we use in FIRST - servos, encoders, camera, and closed loop computer control.
Much like autonomous mode, the robots rarely run longer than 15 seconds in any one program, and many of them require frequent human interaction. Oh - and they have no legs.
Every one of the technologies the kids learn in FIRST are applicable to the Robotics field - from business to cad to the controls sytems and the manufacturing.
robot |ˈrōˌbät; ˈrōbət|
noun
a machine capable of carrying out a complex series of actions automatically, esp. one programmable by a computer.
Merriam-Webster Online -
Definition of ROBOT
1 a : a machine that looks like a human being and performs various complex acts (as walking or talking) of a human being; also : a similar but fictional machine whose lack of capacity for human emotions is often emphasized
b : an efficient insensitive person who functions automatically
2: a device that automatically performs complicated often repetitive tasks
3: a mechanism guided by automatic controls
Based on both of these it appears for the most part we aren’t building “robots” as most teams focus on teleoperation rather than autonomous in my experience. However, given that many of our robots are equipped with senors to provide autonomous function we do have “robots” during the autonomous mode even if they drive in a straight line.
I’ve taken a few robotics courses at CMU and have talked with a few of the professors in the robotics institute. What I stated pretty much reflects their opinion. This is coming from some pretty respected names in the robotics field such as Manuela Veloso and “Red” Whittaker.
I don’t think it is a bad thing that FIRST robots are not really robots. The goal of FIRST is to change our culture into thinking that science and technology are really cool. If the rules changed to ban/greatly reduce teleoperation, then the games will have to be really simplified as making a “smart” AI is really hard. The game would be much less accessible as 1) programming AIs is really frustrating and teams w/o a strong foundation in programming would really struggle to arrive at anything useful and 2) the game will be much slower paced and probably be perceived as boring by the audience watching the game. Compare a FIRST match to a Robocup match using the NAO humanoids. It would be nice though if FIRST gave larger incentive to successfully doing autonomous. I don’t think FIRST will ever be able to go completely autonomous anytime in the near future but if a team is able to go above and beyond in implementing autonomous, why not reward them?
I don’t actually compete in FRC, but I am doing FLL and VRC (and I’m still deciding if doing both in the same season was smart :rolleyes:)
My thoughts:
Are the machines built in FRC really robots during the autonomous period?
From what I’ve read here, I think the consensus would be “yes.”
Are the machines built in FRC really robots during the driver operated period?
Debatable.
What changed?
Only the controller changed: In autonomous, the machine is self-guided. In driver operation, the machine is guided by the human.
Therefore, at what point does everyone agree that a machine turns into a robot?
When the machine is autonomous
Conclusion:
The machines built in FRC are “true/real” robots only during autonomous mode, but in driver operated mode, they are just highly sophisticated, computer aided, remote controlled machines.
FIRST is not “being harmful and preying on the ignorance of the average high school student to what a ‘robot really is.’” In actuality, FIRST is exposing students to the challenge of building a “real” robot (at least for autonomous mode), while still keeping the majority of the challenge “easy” enough for non-programmers. This encourages rookies that “Hey, I can do this too!” but still allows more advanced teams to create increasingly complex programmatic aids for the driver.
Sorry for the long post… I’m normally succinct (engineer-like), but I felt that this deserved a little more time.
I guess the dishwasher in our kitchen is a robot, but the NURC underwater robot notBob sitting in the living room isn’t. But the HERO 2000 next to notBob is. Sometimes.
Let’s not leave out the Web’s wonderful definition source “Wikipedia”.
“A robot is an automatically guided machine which is able to do tasks on its own, almost always due to electronically-programmed instructions. Another common characteristic is that by its appearance or movements, a robot often conveys a sense that it has intent or agency of its own.”](http://en.wikipedia.org/wiki/Robot)
I like to tell my students that a robot is a device, or system, that performs planned tasks based on it’s programming and input from sensors that sense it’s environment, automatically.
A robotic devise, on the other hand, reacts to real-time inputs from humans and it’s sensors and performs tasks based on programming and those inputs.
Now I don’t claim that this is a definition of the two, but I do believe it is a fair representation.
That being said, I believe it is also fair to say that as long as the “robot” functions in autonomous mode, it is both a robot and can function as a robotic device.
Does the fact that this machine is not totally autonomous mean that I have learned less? Is the knowledge that I gained obsolete? Am I less of a person making less of a contribution because the robot is not a robot?
MY answer is no. I say that I will not let this molehill derail anything that I have done or am doing or will do.
What, in the real sense of things, does this thread add that cannot be gained from a water game thread?
I think there are different definitions of robot and robotic and they are evolving. As far as robots only being autonomous what about robotic surgery that is completely controlled by the surgeon?
I think I agree with rtfgnow. This debate isn’t really that important. (Although this is a debate that I have often with some of my friends!)
What I learn in FIRST (and VEX) is applicable to the “real world” and that is really all I care about. Whether I am making a true robot or just programming an R/C car doesn’t matter. What matters is the discipline learned in documenting my work and the mindset of a programmer that I have learned. FIRST has succeeded in its goal of inspiring and recognizing science and technology. Whether or not it uses “real” robots is irrelevant.
