Decided to measure empirically. Turns out if you just drop it, very slowly!
Note: daylight was critical for slow mo video.
First off, rotational inertia was not a negligible component as it turns out. If you drop something side-by-side with the yoyo, the yo-yo falls suppppeerrr slowly. Here is my awesome roomie showing this (using samsung galaxy s9 slow-mo — online says it should be 240 fps)
Look how slow it is compared to a battery!
Second, I think the yoyo only spins about 60 rpm!!! Maybe it can be faster when thrown. But wow, my back-of-envelope calculations was an order of magnitude off. Hah!
The yoyo will be spinning fastest near the bottom of the drop. So we can take this clip and run it through ffmpeg to split out individual frames, then simply count the number of frames per revolution.
Now if the video is originally at 240 fps (no way to confirm this for now), and now I am outputting images at 30 fps. Then, I think each frame is covering a longer chunk of time. That is, of every 8 frames at 240 fps I only select 1 frame to get 30 fps. So 8 frames of 30 fps = 64 frames of 240 fps.
Thus we know that each revolution took 0.26 seconds. (60 sec/min / 0.26 sec/revolution) = 231 rpm. Way less than 3000 rpm!
Will have to repeat in daylight to see what the rpm is like thrown.
I’m still not certain I believe in 300 rpm either, mostly as I’m not sure how much to trust the 240fps spec and my math. Next step I guess is I should do the trivial physics problem… but probably I’ll put work toward making the yoyo happen instead.
(The yoyo is from when I took MIT 2.008. We wanted to make a yoyo that whistled as a spun, this feature didn’t work, but we still had a nice ambigram).
(i’ll sketch out this post and update it as i find the time) — actually will just post in two parts jeez how can i write this much — so —
part 1
i caught covid holy f*k
…or so i thought
amusingly i am working through 6.041 on mit ocw which is the undergrad probability class, and like a perfect complement to the lectures I got a real life lesson on conditional probability
(wrote a section on covid testing terms if you’re unfamiliar)
testing positive
flew home to GA for new years, omciron had already picked up quite a bit and so I wore a proper N95 (i actually didn’t have a KN95 / N95 for the malta international trip) and returned to superstitious practices, in this case a new one, mouthwash. still not eating on planes
got a quickvue test via kaiser permanente drive-through testing appointment that had to turn people away. tested negative on 1/1 (three days after my flight)
had indoor dining in georgia 1/1 (begrudgingly) and freaked myself tf out. had 5-6 ft spacing, 20% capacity (in a place that seats 200 so v. spacious), my table was 5 people, all double and most triple-vaxxed. (i am not sure why it’s more impolite to turn down indoor dining than invite someone in the first place without considering their safety budget…)
dim sum was v. tasty though
i think because of my grudge about being coerced to go (and also… the inviter mentioning having caught covid and having eaten indoors 3-4 times before and not feeling the need to get a booster right away), i was convinced i’d catch omicron because of the dining, like it just felt karma was waiting to get to me after i did my trip to malta and somehow didn’t get COVID (to my parents’ credit they did try to change it to outdoor dining)
i had trouble getting a test in GA (only checked CVS though), and the PCR tents were closed for the holidays, and eventually 4 days passed without anyone having symptoms, so i just flew to boston and planned to quarantine for 5 days and get tested on days 3 and day 5 as before, and also tried to source a day 0 test since it was 5 days out from dining.
roomie also had a sore throat but hadn’t been able to get tested (waiting in rain for half an hour outside a clinic), so great, i could get tests for both of us. i called walgreens close to target (where i was going to curbside pick up some cough drops for him, since i wasn’t sure about spreading airport germs). they had in stock and i picked up four flowflexes with no issue. roommate’s came back negative!
…mine came back positive, wtf??
i felt perfectly fine (zero symptoms) but from what i understood the test was very accurate (cue conditional probability lesson i’ll explain later), like i kept hearing false positive rate of 0.05% so i thought the likelihood i had covid, given a positive antigen and the amount of risk i’d taken over the past week (mostly the indoor restaurant dining and two flights), to be pretty high
i actually thought that, since antigen tests were more likely to give false negatives than false positives, that getting a positive despite having no symptoms meant it was extra likely i had (asymptomatic) covid, even more so than if i had symptoms and had a positive antigen– turns out it’s the exact opposite!
isolation
i freaked the fk out and shut myself in my room right away and asked everyone to mask up around me, and i avoided leaving my room and just went to sleep.
took a second test the next day to convince myself before i really committed to the 10 days of isolation and trying to not spread it to my roomies, and all my roomies had to deal with being a close contact also
also positive, fk me, at this point i started notifying everyone i’d interacted with as i was certain i had covid (certain enough i wasn’t even going to figure out harvard pcr test for confirmation)
which, aside, still have no idea how to notify airlines – i think that comes through dr and department of health? –
presumably you contact a dr and they deal with everything – now i know i guess you can call insurance hotline ! and they will give you to talk to a nurse remotely (nurse hotline) same day
ok i have some trust issues with insurance and the health system, exacerbated by covid-caused instability in living situation, but anyway
=== notes ===
if you haven’t had to do a lot of home testing for covid (e.g. if you are home all the time or get tested regularly for school), here are some terms for you
Antigen test – there are the rapid tests (results in ~15 mins) that are available in drug stores (Walgreens, CVS) for ~$10 to $15
QuickVue, BinaxNow, and Flowflex are the ones I’ve seen in the United States
PCR test – this almost always refers to RT-PCR, reverse-transcriptase polymerase chain reaction. these are the tests that need to be sent to a lab and take 1-3 days for results
Rarer, there’s the isothermal PCR tests, which include Cue tests that Google hands out to its employees, where you take self-sample like the antigen tests but it goes into a cartridge that runs PCR and reports results after 20-30 mins. these are a little more accurate (prolly) than the antigen tests
(Traditional PCR goes through multiple heat-cooling cycles, so I imagine the isothermal part is what saves the time)
Both these are Nucleic Acid Amplication Tests (NAATs), for more see CDC website. The layman’s explanation I came up with for traditional PCR (maybe similar for isothermal PCR?) is you have a search keyword (a primer). if the keyword (nucleic acid) is found, the heat-cooling cycle makes a bunch of copies of it (amplication), which you can then detect more easily (test)
Anterior nasal swab – the PCR and Antigen tests can both use this method which is easy to perform by yourself at home. it’s where you shove a qtip or other stick maybe an inch into your nose and kinda rub it in circles and then (using same stick) do the same for other nostril, which i would’ve found so grosss but now it’s normalized
(other sampled areas include back of throat, or way up in the back of the nose which was what people complained about at the beginning of the pandemic ALMOST TWO YEARS AGO).
Color kit – both MIT and Harvard use these for PCR testing. you self-sample and return to contactless drop off in collection bins around campus. each kit has a barcode which you link to your harvard/mit id online, and results are reported on the color website
COVID facts
Original strain: 5-9 days to infectious/symptoms
Delta: 3-5 days to infections/symptoms
Omicron: 1-3 days to infections/symptoms
Generally can be infectious two days before any symptoms, when you feel healthy!
Asymptomatic = no symptoms, indistinguishable initially from pre-symptomatic
Isolation: stay inside your room, you’re infectious
Quarantine: stay inside your house just in case
Generally now considered that 2-dose vaccination still very effective for preventing severe illness
– so-called “break-through infections”
Booster shots now recommended if you’re 6+ months out
United States topped 1 million cases a day thanks to omicron
In the US, more people died in 2021 than 2020 from COVID
In MA, hospitalizations exceed the peak of last year’s winter surge
For a while it really seemed like Delta might be the last wave, the numbers didn’t go up nearly as much as last year even after all the travelling for thanksgiving. And for a while it was unclear if omicron would actually behave very differently if a large portion of the population was vaccinated. but it sure did !! booster shots have become mandatory for returning to school for Harvard and MIT
spent christmas weekend making this demo of a robot picking up cubes (it doesn’t know which color cube is where beforehand)
Wellll okay this could be a long post but I’m pretty sleep so I’ll just “liveblog” it and not worry too much about making it a tutorial / step-by-step instructions / details. (it’s a two day hackathon project anyway)
two out of three of my roommates are also roboticists, so we (the three of us) started on a house project: a robot arm.
one of us want to focus on designing the arm from scratch (started on cable driven, now focused on cycloidal gearboxes) , and two of us wanted to get started on the software side (divide and conquer). so we got a commercial robot arm for $200: the lewansoul 6dof xarm.
i decided over christmas that it was time to give up on work and do side projects. so i spent two days and hacked together a bare minimum demo of a “helper robot.” we eventually want something similar to dum-E robot in iron man, a robot that can hand you tools etc.
the minimum demo turned out to be verbally asking the robot to pick up a cube of a specific color
the name
i felt mean to call a defenseless robot “hey dummy” all the time, so went with suggestion to use “hey smarty” instead …
probably the dum-e reference will be lost on everyone, since I only remembered jarvis in iron man before starting this project
first print some cubes
T^T the inevitable thing that happens when you get a 3d printer, you crush your meche soul and print cubes. then cover in tape to get different colors
lewansoul robot arm python library
there are several python libraries written, we worked with https://github.com/dmklee/nuro-arm/
which implements already the IK (inverse kinematics) so that we can give x,y,z coordinates instead of joint angles for each of the six joints. runs smoothly (after fixing a udev permissions issue). the library’s calibration required us to move the servo horn a little from how we initially put the robot arm together.
quickstart
$ git clone https://github.com/dmklee/nuro-arm/
$ cd nuro-arm
$ ipython
>>>
from nuro_arm.robot.robot_arm import RobotArm
import numpy as np
robot = RobotArm();
robot.passive_mode() ;
# now robot servos do not fight you. physically move robot arm to desired pose
>>>
test_pose_xyz = np.round(robot.get_hand_pose()[0], decimals=4);
print('---> ', test_pose_xyz, ' <---')
>>>
POS_MID = [0.200, -0.010]
Z_SUPER_HI = 0.10
OPEN = 0.5
robot = RobotArm()
xyz = np.append(POS_MID, Z_SUPER_HI)
robot.move_hand_to(xyz)
robot.set_gripper_state(OPEN)
import easyhid
en = easyhid.Enumeration()
devices = en.find(vid=1155, pid=22352)
print([dev.description() for dev in devices])
# should something like
# devices ['HIDDevice:\n /dev/hidraw2 | 483:5750 | MyUSB_HID | LOBOT | 496D626C3331\n release_number: 513\n usage_page: 26740\n usage: 8293\n interface_number: 0']
make sure robot is plugged into laptop
voice interaction
the other part of the demo is voice interaction. two parts:
speech to text: recognize the trigger phrase and the commands after. this is modelled after the amazon alexa, apple siri, google home style of voice interaction with “AI”, where you say a trigger phrase.
now based on our inspiration it would be “hey dum-e” but this felt pretty mean, so was suggested to use “hey smart-e” instead 🙂
voice interaction part A: generating smart-E’s “voice”
from an earlier project (creating a welcome announcer in the lobby for the gather.town for ICRA 2020) I knew I could use mozilla TTS to generate nice-sounding voice from text.
(as opposed to the default “espeak” which sounds like horrible electronic garble)
was rather annoying install at the time (a year ago?) but seems to be a cleaner install this time around, however!! all the instructions are for training your own model, vs I just wanted to use it to generate audio.
So here’s a quickstart to using Mozilla TTS.
~$ python3 -m venv env && source ./env/bin/activate && which python && pip3 install --upgrade pip && pip3 install wheel && echo done
~$ pip install tts
~$ MODEL="tts_models/en/ljspeech/glow-tts"
~$ tts --text "What would you like me to do?" --model_name $MODEL
~$ tts --text "What would you like me to do?" --out_path "./whatwouldyoulikemetodo.wav"
> tts_models/en/ljspeech/glow-tts is already downloaded.
> Downloading model to /home/nrw/.local/share/tts/vocoder_models--en--ljspeech--multiband-melgan
> Using model: glow_tts
> Vocoder Model: multiband_melgan
> Generator Model: multiband_melgan_generator
> Discriminator Model: melgan_multiscale_discriminator
> Text: What would you like me to do?
> Text splitted to sentences.
['What would you like me to do?']
> Processing time: 0.7528579235076904
> Real-time factor: 0.3542274926029484
> Saving output to whatwouldyoulikemetodo.wav
$ play tts_output.wav
voice interaction part B: giving smart-E “intelligence”
hot stars, in the last year or two we gained a realllllly nice speech2text library. used to be that if you wanted reasonable real-time-ish text transcription, you needed to get a google API key and use google API.
now you can just pip install the Vosk library, download a 50 mb file, and get real-time offline transcription that actually works! and if you want Chinese transcription instead of English, just download another 50 mb model file.
def find_keyword(voice_result):
keywords = ['marty', 'smarty', 'smart']
(In a while True loop -- see test_microphone.py in the vosk repo)
if rec.AcceptWaveform(data):
sentence = rec.Result()
if not state_triggered:
found_keyword = find_keyword(sentence)
if found_keyword:
state_triggered = True
# if state_triggered, then move on to listening for colors
arucotags in python
how to locate the cubes? okay, actually i simplified this step in interest of finishing something in a weekend (before leaving for home for the holidays), and only had three positions. instead the issue was finding which color cube was where.
how to distinguish yellow cube from yellow stickynote? or black cube from checkerboard? i solved it by putting tags on the cubes and sampling (in 2d space) the color slightly above the tags to determine cube color.
note that the gist shows how to get 3d out, vs for the hackdemo i only used the 2d.
what color is the cube
this was actually annoying. in the end i assume there are only three colors and do a hardcoded HSV filter. Specifically, use hue to separate out the yellow, and then value to separate black from blue. (this is hack – note that the logic is flawed, in that black can be any hue)
crop = image[y-rad:y+rad, x-rad:x+rad]
color_sample = crop.mean(axis=0).mean(axis=0)
rgb = color_sample[::-1] # reverse, color is in BGR from opencv
rgb_pct = rgb / 255
hsv_pct = colorsys.rgb_to_hsv(*rgb_pct) # fxn takes 3 values, so unpack list with *
hsv = np.array(hsv_pct) * 255
print('rgb', rgb, 'hsv', hsv)
hue = hsv[0]
val = hsv[2]
closest_color = None
if 10 < hue < 40:
closest_color = 'yellow'
if hue >= 50:
if val > 90:
closest_color = 'blue'
else:
closest_color = 'black'
additional notes
i’ll think about making a public snapshot of the git repo for the demo, perhaps. for the above code samples, my file directory looks like so:
$ ls
env
nuro_arm
speech_output # wav files from mozilla tts
voice_model
hackdemo.py
$ python hackdemo.py
Note
The pop culture reference for smart-E
this is the voice interaction, 38s sec to 62 sec
for longer context of the dum-e scenes, see
creepy voice output
it’s a deep learning model, so it gives unpredictably different results for hey! vs hey!! vs hey!!! and seems to do poorly on short phrases in general. (also you can’t force any specific intonations, like a cheerful “hi there!” vs a sullen “hi there”)
here’s the creepy output of the default model with just “hey”
funny bugs
i got the voice recognition to recognize the colors, but it seemed like the program would get caught in a loop where it would reparse the same audio again and again. i spent a while going “??? how do python queues work, do they not remove items with get()???” and clearing queues and such. eventually my roommie pointed out that, i have it give the response “okay picking up the yellow cube” so it was literally reacting to its own voice in an infinite loop 0:
it was fun to think about the intricacies of all the voice assistants we have nowadays
next steps
on the meche side, build an actual arm that can e.g. lift a drill (or at least a drill battery)
on the cs side, we will continue working with mini arm. bought a set of toy miniature tools (used for kids). we’ll use probably reinforcement learning to learn 1. what pixels correspond to a tool 2. which tool is it 3. where to grasp on the tool. and then 4. how to add new tools to workspace
decisions: match cad models? or pixels? to determine which tool
on the voice side we also have to do some thought about how people will interact with it. “hand me the quarter-inch allen wrench” or “hex wrench” or “hex key” or “t-handle hex” or “ball-end hex” etc., asking for clarification, …
currently we just figured out how to go from webcam image (x,y pixels) to robot frame (x,y,z millimeters) so you can click on the camera image and the robot will move to that point in real life
