Видео

yes

He got into MIT but turned it down???

@@ayyuhhshh2169 no he got in he said for undergrad, but he is now at carnegie mellon for his PhD

I am now a Ph.D. student at the Robotics Institute -- Carnegie Mellon University 🤖

@@yoraish ​ Congrats man.

@@yoraish what was your undergrad degree in?

@@slightlyokvideos Electrical Engineering and Computer Science

It's for a request to get accepted in MIT

It's so MIT can see their potential

your slow ass sure as hell didn’t get in

A slight buzz on your apple watch when you're off pace is much more efficient and only requires the parent hardware.

@@michaelroberts8397 I think a physical object in front of you is better than a vibration. Because adjusting speed while following objects is automatically done by our subconcious mind. Reacting to a vibration isn't. That's why many record holders in marathon disciplines have different people running infront of them for parts of the distance. So this is better adapted to the "software" of the human than a vibration signal. And you can produce them well under 100 bucks, which is very affordable for most track teams.This is my opinion however, but with a small sample comparing perfomance and by asking the runners afterwards what they liked better one could easily assess it.

@@Malaka1802 Nope

It's good for a project but not as a commercial product since it can be easily replicated by chinese companies or anyone else.

Hey Nadav! What high school are you working with? Just curious :) 1. The Raspberry Pi is actually not doing anything in this video. All calculations happen on the onboard flight controller (FCU). The model I am using a Pixracer as my FCU, with ArduPilot Copter 4.1.5 running onboard. An optical flow sensor (ThoneFlow 3901U) and a laser range finder (VL53LX) are connected directly to the FCU. I am thinking of writing a small tutorial soon. 2. I would say that latency is extremely important for holding position. I would try to stay away from offboard calculations of images if I were you. It is definitely possible, but might be computationally expensive. It could be a fun project!

@@yoraish thanks for the reply! I go to a highschool whose name rhymes with zelenskyy :) Does optical flow sensing require a special sensor or could I technically do it with any plain old camera? The reason I'm considering offloading the calculations to a ground control computer is because computing resources are already pretty tight on my custom fc and I doubt I'd manage to add in image processing on top of what I already have. My current problem is that I only have immediate access to an mpu6050, it has no magnetometer so I'll probably get yaw drift while hovering. I was thinking that as long as I have a way to time the delay from offboard optical flow calculations then I can use that to integrate yaw with short term imu readings and correct for the drift. This is my first time doing anything like this so I have no idea if what I'm saying makes any sense.

​@@nadavram7322 Awesome, I have some good friends who went there :). Anyhow, you make a whole lot of sense. Frankly, if you are a student in high school, then you don't only make sense but also have super impressive knowledge! I am no expert, but in general, you can perform optical flow calculations on any image stream. You might, however, get better results with a dedicated optical flow sensor (higher framerates, less compute loads, etc.). I chose to buy cheap optical flow and range sensors and integrate the values from those into my system. I think that this could be a good way to go for you too 🤖🤖. A thing to remember is that you get relative motion between frames via optical flow calculations. This means that you will not have any global heading (yaw) information from optical flow and you'll still be sensitive to yaw drift if you only use data from an IMU and an optical flow sensor. If you look at my video carefully, you'll notice that my drone drifts counterclockwise slowly. If you want to tackle heading drift then you may want to consider integrating information regarding absolute yaw orientation. This could come from a magnetometer (in places that do not have magnetic interference -- outside fields are good) or a camera that tracks a point. Good luck!

@@yoraish thanks again! This helps a lot :)

The secret knock lock is actually not that impressive. It's actually one of the examples in the official Arduino starter kit.

@@weirjwerijrweurhuewhr588 Damn stop hating bruh

@@Itaintme31178 Are you one of those sensitive people who get offended by the truth?

@@weirjwerijrweurhuewhr588 The most sensitive shit I've seen in a while is you taking time out of your day to hate on this dude's achievements. Sounds like someone didn't get in lmao

@@weirjwerijrweurhuewhr588 they werent offended because of truth, irritating part is your lack of communication skills and wording your sentence horribly. using buzzwords like "sensitive" and "offended" will not make you any more correct.

Yeah! So from a high level, Arduinos are a platform used to interface with motors, sensors, etc. And normally do some "simple" computations. Check out this repo for the C++ scripts for the Arduino and Python code for the Raspberry Pi github.com/yoraish/lidar_bot/tree/master/src/arduino . In general, you can read the (somewhat incomplete but perhaps helpful) guide here for more yoraish.com/2020/01/23/a-full-autonomous-stack-a-tutorial-ros-raspberry-pi-arduino-slam/ .

@@yoraish Thank you hopefully when I’m done with everything I’ll share a link here. Then we will meet on the track hahah Thanks :)

Ask away!

I am trying to build something similar to this. I bought a 2WD robot chassis. I want to use an ultrasonic sensor and a camera to help navigate around a basketball court. I’m a 3rd year university student and I’m making a personal robot trainer for basketball but I haven’t really worked on the autonomous part any ideas? And how were you sending commands from the pi to the robot ? I also have the arduino too.

@@geraldmusandirire5446 Sounds awesome! I think the path you'd end up taking would be heavily influenced by your goal. Do you want to follow people on the court? Do you want to follow line? Just do obstacle avoidance? I think the task at hand would dictate your plan of attack :) . In general, I think a good place to start for communication between the Raspberry Pi and an Arduino would be in this repository github.com/yoraish/raspi_bot . I am not sure why I haven't added it to the post. In particular, check out the RaspiBot.cpp and RaspiBot.hpp files, and the wiringSerial library used there. Check out the ball-recognition branch too. If I were you, I'd get the robot running first, then running via commands from the Raspberry Pi, and only then start to work on autonomy. Hope that helps!

@@yoraish Gonna start off by saying thank you like thank you a lot. I was planning on just making it go to certain points on the court where it can watch the user and watch the exercises. So maybe I might need to find a way to get a basketball court map on it? Then maybe I will just need obstacle avoidance of balls and people on the court but it's mainly meant for one/two people really, not a whole team. I included Alexa on it so it can interact with user.

I think both have their pros and cons, but one could never completely out do the other, ideally they should just be used in tandem

Hey! Sorry it took me a while to respond! Unfortunately I did not have a chance to get to that yet... A lot has happened in the world in the last couple of months so I had to put this project on hold. Sorry about that!!

@@yoraish Thankyou for the reply Sir , No issues. 😊

@@Jp_Robotics Check out the tutorial again, updates have been posted :)

@@yoraish Sure Thankyou!

You’re trolling?

Seriously, no shame? It obviously has a lidar unit on the top, it's a four wheel vehicle and it says, lidar, raspberry pi, arduino. Can you not sit down, do the work, read the material, build your damn lab out and work on this if you are interested? This pisses me off to no end. Consider your request? You typed that, you can't google the rest of it? Damn man all of this stuff is wide open technology. If you are interested DO SOME WORK.

@@TheJacklwilliams You should be ashamed of yourself.

Check out the guide at yoraish.com

So go do the research.