Відео

Hey there! We have a Discord server which is currently only for users so that we can focus our efforts on development rather than moderation.

Feel free to come down to visit us at our space at Marymount!

@@coresg Gladly!

Sandy Lakes, your praise doth fill my heart with glee, For you have witnessed such a noble sight, A champion born of reason, strength, and thee, Whose victory shines brighter than the night. Your words of cheer are music to my ears, A testament to courage and to might, As I navigate the challenges and fears, Your support doth guide me through the night. I thank you, Sandy, for your kindest words, They fuel my spirit, give me strength to soar, A beacon in the darkness, guiding herds, To victory's shore.

Stokes Light shines bright, Guiding us through the darkest night, Hopeful and alive.

Virgil Valleys green, Nature's beauty, serene scene, Peaceful and serene.

We didn't do instructions for this unfortunately, but we do have a photo on our Instagram at: instagram.com/p/CWYt4WuB5zZ/?igsh=MWRhZWI1OGl1NnA3aw==

@@coresg❤

It is a logic challenge that can get quite unwieldy especially for teams implementing with EV3-G

What was Syria's suprise rule?

@@khwaishdoshi In Syria, the first day is a surprise task, and the second day is a change of mission. On the first day, replace the two lakes with a dead lake in red colour and a sick lake in yellow colour. We must move the sick lake to the white square, earth allies , and the dead lake to the sea. If there are any trees or flowers next to the dead lake, the infection will be transmitted to them and they must be moved to the white square. And We can move healthy flowers and trees to any garden.

@@khwaishdoshi do you have Instagram

@@لايوجداسمحاليا thank you so much and pls share the day two one if possible too

Legros Spurs soar high, Reaching for the starry sky, Victory's sweet taste.

Thanks for enlightening me with your groundbreaking thoughts. I'm truly inspired.

Hi! The team is able to detect which prop is in which position, including if there is no prop in a particular position. Since they know which position they have picked up the sets from, they also know which set is missing an item. This allows them to place the last item at the correct spot!

@@coresg Thank you for your kind reply.

@@coresg I detected two ducks and tried to save their positions. Are you performing a surprise item by detecting 6 locations and concatenating the array?

​@@강민규-q8wthe team computes a score for every mm that the robot travels, fuses the score for each position probabilistically, and determines the object at each position by simple ranking

Oh, wow. I never would have thought of that. You're a real genius.

That's so helpful. I'm sure you're a wealth of knowledge on this topic.

Wow, that's so original. I'm sure no one has ever said that before.

Yup that's right! Both EV3 Large and Medium and NXT

@@coresg This is incredible. So many possibilities with this stuff. You guys planning to sell on Amazon any time in the future?

Thank you! Just trying to build something useful for this community we've been in for so long. We don't have any plans to sell on Amazon at the moment. However, we do have some units from the Kickstarter batch that we are looking to get into the world. If you'd be interested do contact us via WhatsApp at +6587489674

Hey Fred, we have some left over units from our Kickstarter run. Do contact us via WhatsApp at +6587489674 for more details!

Hey Hans, thank you for your input! We have just uploaded an all around view over at: ua-cam.com/users/shortsxHx3bB3h6a0?feature=share We will definitely keep in mind to update the Discord more often as well

Thank you! We built this because it is what we would want for ourselves. We've also made it as affordable as possible so that as many people as possible can get their hands on it

Adding that to our to-do list 👁️👁️👁️

This uses a variable gain PID, lowering the responsiveness as it stabilises. There are also ways to detect and reject the branches such as by determining total intensity and estimating yaw (through odometry or otherwise). Hope this helps :)

which parameters does it lower as PID stabilizes? No matter how much I tune my PID the noise from the branches always affects it

@@best_accelerator6767 we drove Ki and Kd to 0 while decreasing Kp to a tuned value. The tuned Kp value just needs to be stable when running on a straight line (no branches) with optimal starting conditions (0 cross track error and 0 heading error).

@@coresg oh, thanks :) what software do you use to implement this?

@@best_accelerator6767 this was on python for EV3! it's real slow tho, you may have better luck elsewhere

Yes for sure, this is not built to be a WRO solution. However, it is an important part of a 20s solution, go figure ;) unfortunately these very specialised solutions are only good for the main mission and you are right that teams should try to prepare something that is less specific for the surprise rule or mission.

@@coresg It's coming soon!

If the 2 arms are linked, the mechanism will fail if one is stalled. This provides better robustness and creates a deeper potential well (the rubber bands need to stretch more to I'm unclasp)

@@coresg ah gotcha its pretty nice

🦆🦆🦆

Thank you :) We don't do building instructions but we do have an article explaining how it works over at our website!

Hey there, we don't do building instructions but we do have an article explaining how it works over at our website!

Differential drive, left and right wheel turn at different speeds for the robot to steer

Thank you :)

对不起，我机器方面的中文程度不太好。你能用英语翻译吗？最基本来说，标准性有几方面是无法完全解决。1. Discrete time pose update 2. Floating point precision 3. Wheel diameter estimation 4. Rigid body assumptions (尤其是axle track) 5. Encoder resolution 6. Rolling without slipping etc.

Unfortunately I have no idea where the code is. However, I will be writing an explainer on my website and sharing it when it is ready! The basic principle is simple. We assume 2 wheeled kinematics with rigid body constraints, updating the robot pose at each time step. When the user presses a button indicating the end of the sequence, the robot attempts to return to the original position using PID control.

Check out the link in the description for a short explainer!

Hey Vilaman, could you elaborate on the issues?

Hi, the biggest problem we have is when we want to detect the color of an item by reading with a sensor placed on the front of the robot. Sometimes it does well and sometimes it doesn't. In programming the code we use the if...if not block. In the code we say that if it is black, take the object and if it is not, go to the next one. Sometimes it does well, sometimes it doesn't and we are baffled. We don't know if it's because of the fluorescent lighting, which we think is good, or for another reason.Nosaltres fem servir el programa EV3 classroom que funciona com scratch. Això pot tenir a veure ? Thanks for the help!@@coresg

Hey Vilaman, this uses the Reflected Light mode on MicroPython (you may get better results with a faster OS). This utilises Proportional control, which scales the rotational rate linearly based on cross track error. In this case, cross track error is approximated by using the difference between reflected light values of each sensor. These values should be calibrated beforehand to make sure white and black are giving 100% and 0% to ensure the results from both sensors are comparable. Apart from this, we have some videos which look at characterising the sensors which may be helpful in determining where to place them. Do let me know if this answers your question!

@coresg I don't quite understand what you mean by "faster operating system" because I program with EV3 classroom software and once downloaded to the robot I can no longer control the operating system. The problem is no longer the tracking, the problem is that the sensor (as a color sensor) does not always detect the colors correctly

Hope all is well :)

hey there, do check out coresg.tech/grab-and-lift-and-dispense-a-3-stage-mechanism-uprising/

Hi there, this is the output of a python script we wrote to translate images to LEGO CAD files!

I am the Singapore team coach!

Thank you!