Remote Lab Infrastructure for Distance Learning Courses at the Undergraduate Level in Embedded Systems Design

Author(s):  
Peter Balog ◽  
Martin Horauer ◽  
Peter Ro¨ssler

Lectures and labs with hands-on trainings and project based aspects are of high relevance for courses dedicated to embedded systems design when the transfer of practical skills is a major objective. Therefore, small classes, good support and frequent access to target platforms over a long period of time are beneficial. Providing access to the latter, however, can become a hassle for the organizers, especially when multiple courses are to be held in parallel or when they are organized as asynchronous distance learning courses. The problem is even aggravated when the platforms should be kept at the fore-front of the state-of-the-art. For this purpose we present in this paper concepts and implementation guidelines of a remote lab infrastructure that addresses these issues. In particular, the presented approach keeps the efforts to migrate to new embedded targets simple, enables nearly 24/7 times of access, ideally complements on-site trainings, and keeps the required costs low.

Author(s):  
Peter Balog ◽  
Michael Kramer ◽  
Roman Beneder ◽  
Philipp Brejcha

According to labor market needs even fresh graduates from undergraduate programs have to have not only profound knowledge but also extensive practical experiences especially when it comes to software design for Embedded Computing Systems. Didactic approaches like problem-based learning and project-based learning with a high degree of hands-on training using state of the art hardware, software, and tools have proven to achieve this learning outcome. Even though hands-on training using industry relevant equipment works fine with full-time students, the desired practical skills have to be obtained in a different way with part-time or distance-learning students. This is because of the significantly reduced training hours at dedicated university labs. This paper focuses primarily on a concrete setup and mix of dedicated learning infrastructures (“Remote-Lab” and “Hardware/Software Co-Simulation”) suitable for courses dealing with Embedded Systems (Software) Design to support students, participating in part-time or distance-learning degree programs, in developing their required skills.


Author(s):  
Hamid Mahmoodi ◽  
Arturo Montoya ◽  
Joie du Franco ◽  
Chris Rodriguez ◽  
Jose Carrillo ◽  
...  

Author(s):  
Yassine Larbaoui ◽  
Ahmed Naddami ◽  
Ahmed Fahli

This paper presents the work of adapting hands-on laboratory’s materials of NI Elvis and Quanser from in-poste exploit to online access and remote exploit for online experimenting, after analyzing different aspects of adapting any hands-on laboratory’s material of in-place experimenting to remote exploit. This paper presents the work of developing a software multiplexing technique, and other techniques, to multiplex between different software codes and programs, in order to control different types of experiments in electronic of energy while using and sharing the same physical components and materials nearly simultaneously. In addition, this paper presents the work of creating web client interfaces; to use those embedded systems of NI Elvis and Quanser and their deployed experiments through the internet while relying on an e-learning platform of our remote lab to support their remote access. The principal advantage of conducted adaptations is sharing the same hardware and software resources between different experiments at the same time, while exploiting them locally and through the internet by multiusers.


Author(s):  
Ken Ferens

This paper reports on a project based learning approach taken to teach the ECE 3740 Systems Engineering Principles and ECE 3730 Principles ofEmbedded Systems Design courses at the University of Manitoba. These courses were 100% hands-on, and each student was given development hardware and software in a lunch box to take home and work on projects throughout the course. Industry representative projects were chosen based on the author’s 5 years of experience working in the embedded systems industry. The courses were given in a company-like setting, where the lectures and laboratories were organized as product requirements gathering and analysis, design modeling and review, test plan and procedures, engineering change request and management, documentation, and product deployment meetings and events. The test and final exam were performed by students in the laboratory; they brought their embedded systems hardware in the lunch box,solved the given hands-on problems of the test/exam, and demonstrated their solutions, in real-time. This novel methodology allowed the examiner to directly assess student performance in the CEAB attributes of Design, Analysis, Investigation, and Tools, because their designsand solutions were actually demonstrated in actual hardware and software, not just on paper, like the conventional assessment approach for tests and exams.


2021 ◽  
Vol 11 (3) ◽  
pp. 1093
Author(s):  
Jeonghyun Lee ◽  
Sangkyun Lee

Convolutional neural networks (CNNs) have achieved tremendous success in solving complex classification problems. Motivated by this success, there have been proposed various compression methods for downsizing the CNNs to deploy them on resource-constrained embedded systems. However, a new type of vulnerability of compressed CNNs known as the adversarial examples has been discovered recently, which is critical for security-sensitive systems because the adversarial examples can cause malfunction of CNNs and can be crafted easily in many cases. In this paper, we proposed a compression framework to produce compressed CNNs robust against such adversarial examples. To achieve the goal, our framework uses both pruning and knowledge distillation with adversarial training. We formulate our framework as an optimization problem and provide a solution algorithm based on the proximal gradient method, which is more memory-efficient than the popular ADMM-based compression approaches. In experiments, we show that our framework can improve the trade-off between adversarial robustness and compression rate compared to the existing state-of-the-art adversarial pruning approach.


Sign in / Sign up

Export Citation Format

Share Document