Hacker News Reader

Comment by protonfish | Hacker News Reader

I am starting to get annoyed by the overloaded term "robot." This article is really about automated warehouses. That is interesting, but is sufficiently advanced automation truly robotics? If so, where does it cross that line? Many hobbyist "robotics" (think BattleBots) are just remote-controlled vehicles. Neither have autonomous behavior. It'd be nice to have terms for each but I can't think of what would be appropriate.

View on HN

I teach robotics to students age 11 - 14. We have settled on the definition that a robot must do three things: sense, think, act. First thing I do is take away all of the remote controls and tell them they have to program it.

I wish that programs like VEX and FLL would do away with the "tele-operated" portions of their competitions. Kids may enjoy it, but too many of them fall into the trap of treating it all like a remote controlled toy and never get very deep into the programming aspect.

How and where did you get this job? It sounds like my dream job

In Chicago there's a robot store always looking for teachers to help students about that age range (maybe even younger). Can't remember details, but if your city has a robotics store, chances are they can help you get into teaching.

I don't remember FLL having a teleop mode. Back when I did it, it was fully autonomous, but things may have changed since then.

However, FTC, supposedly the grown up version of FLL, had both teleop and autonomous, and with the way it was scored, all you really needed was a solid teleop to crush regional competitions, so we ended up doing way less programming than in middle school FLL. I believe VEX worked the same way, but I only did it one year, so my memory of it is fuzzy.

Isn’t it FRC (FIRST Robotics Competition)?

It’s kind of simplistic to say all you need is a good teleop period. In many competitions some teams that had worse teleop were picked because they could consistently score in autonomous, which is pretty good EV.

Having worked on some warehouse automation equipment briefly in the past, I'd say they can meet that definition. The equipment I worked on was called a tilt-tray sorter system.

Packages would arrive along a conveyor belt and get put on a tray in a big train of trays (like a roller coaster for boxes).

A scanner would read a bar code on the package (sense).

The bar code would correspond to a particular loading bay where a truck was waiting that the package needed to be put on. (think - match code with loading bay).

When the tray went around and reached the chute for that loading bay it would tilt up and slide the box down to where it could be loaded (act).

This makes an Alexa device a robot, especially when integrated in a smart house. I don't like that. (It's an occasional source of disagreement I have with a friend.) I think in order to be a robot, it needs movement. Either it can move something that could reasonably be called "itself", or it has direct peripherals like an arm on the same "itself" that can move things around it. Thus giving an Alexa device a robot arm, or some robot wheels, and letting it move itself or manipulate things around itself, would make it a robot, even if most of the brains are in the cloud.

Of course even with the requirement of mechanical motion, you could still argue a toaster is a robot. It's hard to define robot: http://robohub.org/robohub-roundtable-why-is-it-so-difficult...

I hope this leads to the revival the use of the truly awesome word "automaton" for autonomous robots!

Cool. I would have preferred "android" but Google's smartphones ruined that.

I completely agree the word "robot" is entirely overused in inappropriate ways. However, I evaluated the Kiva system shortly before Amazon acquired the company and can say their system meets your definition and, IMHO, is as robotic as a Roomba.

Seconded -- while the decision-making ability is centralized (not in each individual robot) the system as a whole is still able to make its own decisions in many areas.

My definition of a 'robot' is a machine with motor functions and can make its own decision on how to act in reaction to its environment.

For example, a phone isn't a robot, not unless you give it wheels and the ability to act.

By that definition, a clothes dryer with a humidity sensor is a robot. (And I'm fine with that!)

My favourite robot is a dishwasher.

Mostly because it is so different from what some people naively imagine when you asked them to think of a design for a robot to wash dishes.

I have to try that question on people sometime. Sounds like fun.

Also, reminds me of this:

   Status: Alive

http://futurama.wikia.com/wiki/Amana

The dishwasher is missing the thinking part... It only ever does the exact thing ordered to it.

Bashing one's head against the wall in a day long debugging session for some hard bug usually cures people of the opinion that "computers only ever do what we tell them to do" is a serious limitation on their range of behaviours in practice.

The combinations on my dishwasher can fit in a 12 line lookup table.

A phone isn't a robot

Ahem: https://youtu.be/AKCpqqhe-S4?t=1m

Any phone with a vibration motor would technically be a robot (vibrate when message received)

Since robotics are almost a field of AI it seems like it has inherited the premise that once it has been completed or in production it ceases to be called robotics .

This question/doubt has very interesting consequences.

An argument against the thesis that "robots" will lead to catastrophic loss of jobs, is that the concept of "robot" in a strict sense (intended as a sudden similarity to human intelligence/capacity) doesn't really exist, because "robots" are simply a form automation - and civilization has been incrementally developing automation for 250 years by now.

edit :

sorry, this was what I was looking for,

https://www.theincomparable.com/robot/

In South Africa traffic lights are called robots.

Many traffic lights around the world use buried loops of wire as Hall effect sensors (for example, to register when a car is waiting to make a left turn). So that would pass the sense-think-act test, I'd say, even without considering Smart Traffic Lights[1].

1: https://en.wikipedia.org/wiki/Smart_traffic_light

Reply on news.ycombinator.com