No video

Reachy first prototype (2017)

  Рет қаралды 14,353

Pollen Robotics

Pollen Robotics

Күн бұрын

Reachy is a bio-inspired robotic prothesis arm designed by Pollen Robotics for INCIA research lab.
Its goal is to explore different ways for human-robot control interfaces, in particular, using EMG.
Learn more about reachy: www.pollen-rob...
All sources are available here : github.com/pol...
Learn more about INCIA : www.incia.u-bor...
Facebook : / pollenrobotics
Twitter: / pollenrobotics
Instagram : / pollenrobotics
Music from ous friends : The Geek X VRV, With you (BTOS VOL 3) : • The Geek x vrv - With ...

Пікірлер: 24
@DwAboutItManFr
@DwAboutItManFr 7 жыл бұрын
Howard Holowitz would be proud.
@Nolifounai
@Nolifounai 7 жыл бұрын
Très sympa comme projet !
@nagalakshmiduvvuri2316
@nagalakshmiduvvuri2316 3 жыл бұрын
good one
@williamhuang5329
@williamhuang5329 2 жыл бұрын
Hanzhen harmonic drive gear , strain wave reducer, robot joint , over 30 years experience
@sachanaoun1573
@sachanaoun1573 4 жыл бұрын
J'adore votre idée, keep on !
@PollenRobotics
@PollenRobotics 4 жыл бұрын
alors tu vas vraiment aimer ce qu'on va annoncer dans quelques jours ;-)
@AntonioLemos66
@AntonioLemos66 7 жыл бұрын
Tenho algumas idéias para fazer uma programação orientada a instintos. Pensei numa forma de podermos programar instintos e fazer com que robôs tivessem vontade propria, procurando atender a seus desejos instintivos pre programados. A idéia seria criar um algoritmo onde os vários instintos fossem aumentados em sua prioridade com o tempo em pesos diferentes, e o robô procurasse então priorizar o instinto A ou B com alguma tarefa a ser realizada, se o instinto tivesse uma maior pontuação de prioridade. Ao realizar a tarefa a prioridade do instinto atendido seria diminuida. Ele iria constantemente analisar o meio tentando indentificar objetos e as formas possíveis de se atender ao instinto A ou B. Suas ações iriam então a depender do meio, de seus instintos e de sua análise do ambiente em seu mapa mental digital, procurando atender ao instinto de maior prioridade da vez, uma análise posterior iria fazer com que ele percebesse o meio e identificasse se atendeu ou não o instinto com os passos auto programados, se não, iria procurar atender outro instinto ou procurar um novo sequenciamento de tarefas, possíveis para assim tentar novamente a tender o instinto prioritário. Para sequenciar ações com o objetivo de atender seus intintos, ele iria fazer varias simulações em seu mapa mental , em um processo de tentativa e erro e seleção natural dos sequenciamentos de ações que funcionam ( aprendizado / criatividade ). Com esta metodologia de programação iria ser possivel criar instintos conflitantes e fazer com que o robô tivesse uma complexidade maior de tomada de decisão que iria parecer que ele tem vontade própria, como ocorre com o ser humano, porque iria agir de formas inesperadas conforme o meio que encontrasse e percebesse com seus sensores e conforme o que aprendeu em suas experiências, que seria a comparação do que foi planejado e o resultado obtidos percebido. exemplo de instintos simples: - buscar pela luz - analisar objetos e classifica los - curiosidade, tentar mapear outras áreas. - medo , se afastar de objetos desconhecidos que se movam - fome , busca por energia - dor , fuga de ambientes nocivos ao sistema, temperatura, frio, sobre tensão dos motores - amor , identificar objetos/sujeitos que deve cuidar e atender ordens - confiança, tendencia de repetir o que funciona ou ter menos medo de sujeitos novos identificados que não oferecem perigo com o tempo. entre vários outros institos que podem vir a ser similares aos humanos ou totalmente diferentes. O robo iria mapear espaços e classificar objetos, sujeitos para o robo, seriam objetos capazes de se mover ou realizar alterações no espaço percebido.
@joshkar24
@joshkar24 2 жыл бұрын
can this be controlled with a small version of the arm as a controller, or even a version that you strap over your own arm, perhaps with force feedback ? the precision could not matter so much provided a visual-tactile loop was constructed. Obvious sci-fi fantasy, but I can't wait for such tech, so much potential !
@damien648
@damien648 6 жыл бұрын
Excellent projet, bravo pour votre travail et merci à INCIA. Je contribue aussi au développement d'un bras humanoïde 5 actionneurs (sans le poignée), notamment en ce qui concerne l'hardware et la mécanique. J'aurais beaucoup aimé savoir pourquoi l'axe 1 est horizontal et non vertical, ce qui entraîne plus de couple (je suppose que c'est pour le rendre plus humain). De même j'utilise aussi des servomoteurs de la gamme MX et je me demandais pourquoi vous aviez décidé de ne pas contre balancer les axes, et du coup je voulais savoir qu'elle était la masse que pouvais porter le bras et qu'elles étaient l’accélération maximal en bout de bras. Ps: Où étiez-vous à Vivatech 2018 ? :c
@MatthieuLapeyre
@MatthieuLapeyre 6 жыл бұрын
Ça fait beaucoup de questions tout ça :) Globalement le bras a été conçu pour reproduire les mouvements du bras humain pour un labo qui fait de la recherche sur l utilisation de prothèses. Donc on a fait nos optimisation dans ce sens. Il y a en effet beaucoup d'autres choix de conception possible et bien meilleurs pour d'autres cas d'applications. En particulier si on veut faire de la manipulation.
@NicolasGonot
@NicolasGonot 7 жыл бұрын
Joli projet et jolie vidéo :) Par contre le lien vers les sources de marche pas :'(
@MatthieuLapeyre
@MatthieuLapeyre 7 жыл бұрын
Merci, c'est corrigé !
@khenafiabderezak2563
@khenafiabderezak2563 6 жыл бұрын
Bonjour toujours Probleme de se connecter au page source :(
@AntonioLemos66
@AntonioLemos66 7 жыл бұрын
I have some ideas for instinctual programming. I thought of a way we could program instincts and make robots have their own will, trying to meet their pre-programmed instinctual desires. The idea would be to create an algorithm where the various instincts were increased in priority over time in different weights, and the robot would then try to prioritize instinct A or B with some task to be performed if instinct had a higher priority score. When performing the task the priority of the attended instinct would be diminished. He would constantly analyze the medium trying to identify objects and possible ways of attending to instinct A or B. His actions would then depend on the environment, his instincts and his analysis of the environment on his digital mental map, trying to meet the highest priority instinct of the time, a later analysis would make him perceive the medium and identify whether it met or Not the instinct with the programmed auto steps, if not, it would seek to attend another instinct or to look for a new sequencing of tasks, possible to thus try again to tend the priority instinct. In order to sequence actions in order to serve his subjects, he would make several simulations on his mental map, in a process of trial and error and natural selection of action sequencing that works (learning / creativity). With this programming methodology it would be possible to create conflicting instincts and make the robot have a greater complexity of decision making that would seem to have a will of its own, as it happens with the human being, because it would act in unexpected ways according to the medium That he would find and perceive with his sensors and according to what he learned in his experiments, that would be the comparison of what was planned and the result obtained perceived. Example of simple instincts: - search for light - analyze objects and classify - curiosity, Try to map other areas. - fear, move away from unknown objects that move - hunger, energy search - pain, leakage of harmful environments to the system, temperature, cold, over voltage of the motors - love, identify objects / subjects that must care and meet orders - confidence, a tendency to repeat what works, or to be less afraid of identified new subjects who do not pose a hazard over time. Among several other instits that may be similar to humans or totally different. Theft would map spaces and classify objects, subject to theft, would be objects capable of moving or making changes in perceived space.
@alonmark7565
@alonmark7565 4 жыл бұрын
Brasilian
@jmg8856
@jmg8856 7 жыл бұрын
Combien d'actionneurs? Et on peut avoir une idée du prix? :)
@PollenRobotics
@PollenRobotics 7 жыл бұрын
Il y a 8 actionneurs, les plus gros coutent environ 500 euros et les plus petits 20. Il y a ensuite pas mal de boulot en impression 3D, assemblage, etc ...
@Ayxn_07
@Ayxn_07 4 жыл бұрын
Hey sir,even i am making something similar sir pls i need the 3D printing files of your this project...
@PollenRobotics
@PollenRobotics 4 жыл бұрын
You can download those files here cad.onshape.com/documents/66388ae9c63cef53d76acd77/w/68c2411483d5bc65c7f54234/e/931a3985934243d3ff04a322
@sayyedshadab5516
@sayyedshadab5516 4 жыл бұрын
Sir what type motors are used can you please help me with information of motors and the codes of raspberry pie
@jaquanthomas5595
@jaquanthomas5595 6 жыл бұрын
Im a 15 year old in high school as a freshmond and i want to be a welder engineer that creates the impossible things right now a college uk wants me yo join them and be their for 6 week to help me make the robotic arm for people who don't have a arm can u help me
@PollenRobotics
@PollenRobotics 6 жыл бұрын
Hi, Great news for you! You can find the sources of Reachy on github : github.com/pollen/reachy
@Tubernameu123
@Tubernameu123 6 жыл бұрын
@Dr.tech-
@Dr.tech- 7 жыл бұрын
in 2017 This could be either
「Unitree G1」プロトタイプから量産機へ、驚きの身体能力
1:08
Making Chat (ro)Bots
8:28
Boston Dynamics
Рет қаралды 847 М.
Joker can't swim!#joker #shorts
00:46
Untitled Joker
Рет қаралды 40 МЛН
Yum 😋 cotton candy 🍭
00:18
Nadir Show
Рет қаралды 7 МЛН
Кадр сыртындағы қызықтар | Келінжан
00:16
You've never seen the Robot Joint like this one!
11:24
Skyentific
Рет қаралды 218 М.
The coolest robot I've ever built!
19:40
Thomas Burns
Рет қаралды 4,6 МЛН
Introducing Reachy the new open source interactive robot - 2020
3:04
Pollen Robotics
Рет қаралды 58 М.
New AI Humanoid Robot by Boston Dynamics Just Blew Everyone Away!
5:11
The Robot-Arm Prosthetic Controlled by Thought
5:05
Bloomberg Originals
Рет қаралды 3 МЛН
The Dingo | A Low Cost, Open-Source Robot Quadruped
8:09
Nathan Ferguson
Рет қаралды 237 М.
Human Robot Mass Production Process with New 3D Printer Factory in Korea
10:13
All process of world
Рет қаралды 60 МЛН
How Far is Too Far? | The Age of A.I.
34:40
YouTube Originals
Рет қаралды 62 МЛН
Meet Beomni: Your new robot avatar
8:05
CNET
Рет қаралды 42 М.
Joker can't swim!#joker #shorts
00:46
Untitled Joker
Рет қаралды 40 МЛН