PhD Position: Machine Learning/ pattern recognition of bionic upper-limb devices in activities of daily living

The goal of upper-limb prostheses is to restore lost functionalities and autonomy in daily living tasks, work and social activities. The commercial solutions range from anything between simplistic body-powered hooks to advanced multiarticulated myoelectrically controlled hands. Most of the latter are designed to match the appearance and functionalities of the human hand. Such hands enables one to perform several hand motions through coordinated activation of independently motorized fingers.

These dexterous bionic devices usually use a pair of surface electromyography (sEMG) sensors to control one degree of freedom (DOF) at a time and adopt switching techniques via muscle co-activation, a mobile application and short-range proximity sensors to select the required grasp sequence. However, many consider such sophisticated myoelectric devices unreliable and difficult to use bringing out large variablity between the solutions available and users’ day-to-day needs. Applications are invited for a PhD student to work on the development and application of machine learning/ pattern recognition methods for better control and use of upper-limb prosthetic device in activities of daily living.

The candidate will closely work with Dr. Amartya Ganguly under the supervision of Prof. Alois Christian Knoll (The chair of Robotics, Artificial Intelligence and Real-Time systems) and Prof. Sami Haddadin (The Munich Institute of Robotics and Machine Intelligence (MIRMI) as one of TUM’s Integrative Research Institutes (IRI).

The candidate will be expected to:

    Design, develop and modify ANN frameworks
    Plan and conduct experiments, both individually and in collaboration with clinical experts
    Acquire sEMG, motion capture and other related data.
    Process and analyse signals in realtime
    Generate computational models for simulation in loop evaluation and validation processes.

This work will contribute to a better use of myoelectric prosthesis among people with limb difference, development of new prosthetic devices driven by AI and regulatory procedures for such devices.

Interested applicants should send the necessary documents via email to applications@mirmi.tum.de quoting “PhD in AI driven upper limb prosthesis” in the e-mail subject line. Only shortlisted candidates will be notified. Preference will be given to applications received before January 10th, 2023 as a strict deadline.


More: https://portal.mytum.de/jobs/wissenschaftler/NewsArticle_20221205_152920

RAICAM aims to train a cohort of 10 PhD students to work on the next generation of mobile robots for inspection and maintenance of industrial facilities. RAICAM will develop a multi-domain, multi-agent robotic sample retrieval system that will be able to obtain surface swabbing samples of critical infrastructure buildings for chemical or biological analysis.

TU Munich position: Online planning for bimannual mobile manipulation

Essential Skills & Qualifications

    An MSc qualification in Robotics, Computer Science,  Electrical & Electronic Engineering, Mechatronic & Mechanical Engineering, or other equivalent discipline
    An MSc dissertation on a robotic topic
    Excellent programming skills
    Evidence of practical experimental work in robotics
    Evidence of strong mathematical background
    Evidence of solid knowledge in manipulator modelling and control

Desirable Skills & Qualifications

    Evidence of using ROS/ROS2
    Knowledge/experience relating to one or multiple of the following topics: motion planning, force/impedance control, robot learning, bimanual/multi-arm manipulation.


All positions/More Details: https://www.raicam.eu/#Job_Positions

Description

The Munich Institute of Robotics and Machine Intelligence (MIRMI) at the Technical University of Munich is offering a PhD position focusing on using collective learning for tactile robots to learn manipulation skills to perform various tasks in future factory environments and beyond. We are seeking an outstanding applicant with passion for robotics and AI as well as a strong desire to shape the future AI-enabled, robotized factories. The successful candidate will contribute to the AI.Factory project (https://www.mirmi.tum.de/mirmi/innovation-sectors/work/kifabrik/), which is a flagship project of the Bavaria High-tech Agenda to build the first prototype AI factory in Bavaria and enable research and development in cutting-edge technologies in robotics, telepresence, wearable electronics and AI to revolutionize the future of work in production.

This project will give you a unique opportunity to work with large scale networked robot systems to accelerate learning of robotic skills for manufacturing and logistics processes. You will work closely with researchers at MIRMI from various chairs and industry partners in an interdisciplinary environment to share expertise and collaborate on research publications in top conferences and journals in robotics and AI.

The goal of the first prototype of AI Factory is to show how state-of-the-art IT and mechatronic high-tech components (e.g. computer systems, sensors, mechatronic components, entertainment electronics) can be manufactured economically in Bavaria in the future, regardless of the number of production units. The AI Factory will be able to reconfigure its modular facilities and robots involved for different products and enable Production-as-a-Service. To achieve this, the successful applicant will work on open problems about autonomous learning of robotic skills, involving physical interaction with environments for challenging manipulation tasks, understanding and reasoning task sequences and collaborative AI for robotics. Knowledge and experience will be shared among the networked group of robots to accelerate learning and acquiring of work techniques.

Key responsibilities/tasks

    Contribute to the design and construction of reconfigurable and self-organized AI factory
    Contribute to the design and implementation of networking mechanism of AI robots to enable knowledge transfer for collective learning
    Collaborate with industry partners to generate analysis on skill primitives and taxonomy based on industrial use cases
    Conduct research in autonomous skill learning using learning-from-demonstration / observation, reinforcement learning, genetic algorithms and or black-box optimization
    Conduct research in collective learning to accelerate skill learning using large-scale networked AI robot systems


more -> https://portal.mytum.de/jobs/wissenschaftler/NewsArticle_20220217_202046

Στη βιβλιοθήκη που πήγα σήμερα μου ειπε να παω εκει με στικακι. Έτσι και αλλιώς θα χρειαστεί να πας για να υπογράψει τη δήλωση που χρειάζεται για την ορκωμοσία.

Αντε ρε Λιτσαρδάκη θα βγαλεις κανενα υλικό να παρουμε κανά πτυχιο ???  

Είχε πάει κανένας στο τελευταίο μάθημα που έκανε ασκήσεις ;

Και γω το ίδιο θέμα μέχρι που έκανα και την (decrypt-and-identify) και μου έβαλε 5/10... Ότι να ναι είναι το σύστημα σε αυτήν την εργασία.. Αν σου βγαίνουν σωστά μη μασάς προχώρα.

Έχω κολήσει στο Α ερώτημα... Γράφω την υλοποίηση όπως νομίζω πως είναι και παίρνω ένα πολύ περίεργο αποτέλεσμα..


Κάποιες συλλαβές τις βγάζει σωστά ενώ κάποιες άλλες όχι..   μα πως γένεν αυτό ; έχει κανένας παρόμοιο πρόβλημα  

Thanks guys! αφού έλυσα βλακείες κολλήματα, η εργασία βγήκε γρήγορα.. Εχει φουλ ανούσια κολλήματα λόγω της tester.
Συνοψίζω  το πως δούλεψα αποφεύγοντας τα κολλήματα λόγω tester, πάνω κάτω έχουν ήδη ειπωθεί.
Όσο δούλευα στο αρχείο ps1, είχα ορισμένη την γλώσσα (#lang racket) για να εμφανίζονται σωστά τα συντακτικά λάθη, και να μπορώ να κάνω run το πρόγραμμα. Επίσης για ορθό testing ορίζω τις μεταβλητές time-100 και time-200 με τους κατάλληλους χρόνους (9.58 και 19.19 αντίστοιχα).
Όταν είναι να τρέξω την tester ή να στείλω το αρχείο online σβήνω από το ps1 τα defines των time-100 και time-200, καθως και τη γραμμή #lang racket, ώστε να μην υπάρχει conflict..
Για debuging μπορείτε απλά να γράψετε το όνομα της μεταβλητής στο prompt και να δείτε την τιμή της, οπότε συγκρίνοντας τις τιμές που βγάζετε με αυτές που υπάρχουν στην tester μπορείτε να βρείτε γιατί είστε λάθος.
Η tester για να μπορεί να τρέξει, πρέπει να είναι ορθά ορισμένες οι μεταβλητές ελέγχου, οπότε για τα ερωτήματα που δεν έχετε κάνει ακόμα απλά αντικαταστείστε το με έναν αριθμό, προσωρινά, ώστε να δείτε πως πάτε.

Καλή συνέχεια

Όταν τρέχω την tester, μου εμφανίζει αυτό το error ...

Εχει κανεις ιδέα τι παίζει ;

Που είναι το γραφείο του κύριου διδάσκοντα ;

έχει πέσει η psychee ? Γιατί εμένα δεν μου την ανοίγει

updated: Ολα οκ τώρα !

Μπορεί σήμερα να πάρει κάποιος θερμοκρασίες ;
Έχω δοκιμάσει όλα τα TXX, απο Τ00 μέχρι Τ99 και δεν έχω πάρει τίποτα :/

Ξέρει κανείς τα bonus ανάλογα με την ημερομηνία παράδοσης ;

Είναι στην σελίδα του Πεδίου 2