Weekly Robotics Newsletter Archive [4/27]

  • Weekly Robotics #87

    Issue 87

    The most clicked last week was the Soft Bank’s robot simulator with 14.5% opens.

    1) Virtual DonkeyCar (and Other Cars, Too) Race.
    Meetup
    INFO: Folks at DonkeyCar are hosting a virtual deep-learning fueled competition this week. If you look at the Meetup page you will find a link to a YouTube video showing the simulator in action.

    2) Hello (Subterranean) World!
    GitHub
    INFO: Michael from Open Robotics had shared a tutorial how to get started with SubT virtual challenge. Time for some cave exploration!

    3) SubT Challenge Urban Circuit Overview.
    ROS Discourse
    INFO: While we are on the topic of SubT challenge and Open Robotics: this post by Katherine Scott is the best write up on the teams and robots taking part in the challenge I’ve seen to date. I like how Katherine provides lots of background information and covers the highlights of each of the presented platforms.

    4) iRobot Education Launches to Train Future Roboticists.
    The Robot Report
    INFO: “iRobot has created a new division, iRobot Education, to house all of its education products and services under one roof. iRobot Education includes the Root coding robot, Create 2, a new iRobot Coding platform, and the company’s STEM outreach program.”

    5) Debugging a Live Saturn V.
    zamiang.com
    INFO: Not very much robotics related but this might spark your imagination: imagine approaching a 6.5M pounds (2,950,000 kg) steel barrel carrying 203,400 gallons (770,000 liters) of explosives that is ready to be launched into space just to replace one component. I can see why safety helmets wouldn’t feel like they improve your safety in any way.

    6) How Boston Dynamics’ Spot Robot Works!
    YouTube (Adam Savage’s Tested)
    INFO: In this episode of Tested Adam Savage goes over some of the functions of the Boston Dynamic’s Spot quadruped robots. What I like about videos like this is seeing the system’s from the user’s perspective and finding out about all the features the robots are shipped with.

    7) Access High-Resolution Ocean Data Sets.
    Saildrone
    INFO: Yet another example of using robots for good. At data.saildrone.com you will find datasets captured by these unmanned surface vehicles. Information about Saildrone was previously shared in this newsletter in the issue #63 and #28.

    1) Humble Book Bundle: Artificial Intelligence & Machine Learning by Morgan & Claypool.
    HumbleBundle
    INFO: “We’ve teamed up with Morgan & Claypool for our newest bundle! Get ebooks like Essentials of Game Theory, Natural Language Processing for Social Media, Creating Autonomous Vehicle Systems, and more” - as always with these book bundles if you decide to purchase it then you can choose to support this newsletter and make it better.

    Job Seekers / Careers

    In the issue #83 I’ve started this section to try to help out those looking for work in the times of pandemic. If you are currently looking for work then feel free to send me. Similarly if your company is hiring Roboticists feel free to let me know and I will publish your open positions.

  • Weekly Robotics #86

    Issue 86

    If you are wondering how Weekly Robotics is doing subscriber-wise then you might like to take a look at the Q1 2020 report. In the past 4 months there was a 10% growth in subscriptions, while website visits were down. The most clicked last week were software developer’s thoughts on his low cost robot vacuum with 14.9% opens.

    1) qiBullet.
    SoftBank Robotics
    INFO: qiBullet is a bullet-based simulator for SoftBank robots, licenced under Apache 2.0. Looking at the tutorials it makes the control of the simulated robots easy using the Python API. The simulator can be also interfaced with ROS. Many thanks to Maxime for letting me know about this project!

    2) Industrial Robot Given New Life And Controller.
    Hackaday
    INFO: There are two interesting parts to this article. The first one is the project itself where Hackaday user caltadaniel managed to reverse engineer a Yaskawa motoman MH5LF industrial arm and creating a custom control box for it. The second interesting bit is the video where the robot flicks a switch of an useless box - I found this one quite entertaining.

    3) What is FOC? (Field Oriented Control) And why you should use it! || BLDC Motor.
    YouTube (GreatScott!)
    INFO: If you want to know more about Field Oriented Control for electric motors then this video by GreatScott! is a great introduction to the topic.

    4) Blast From the Past: UBR-1.
    Show Us Your Sensors
    INFO: In this blog post Michael Ferguson, a Founder of Unbounded Robotics and Vanadium Labs and a former CTO at Fetch Robotics, describes some of the features of UBR-1, an alternative to the Willow Garage’s PR2 robot for R&D. For the National Robotics Week Michael has written a series of posts on his blog, if you are into ROS and robots then you might want to check them out.

    5) Building a Robotic Spider.
    Medium
    INFO: Bobbie is an 8-legged spider robot being developed by Adam Purdie. The robot will use Dynamixel servos as actuators and Jetson Nano for the brains. This post is a build log of the whole platform including wiring. I’ll be looking forward to following the project updates, especially that the author plans to run ROS2 on it, make it walk on uneven surfaces, and perform some machine learning using a simulator.

    6) Combating COVID-19—The Role of Robotics in Managing Public Health and Infectious Diseases.
    ScienceRobotics
    INFO: How can robots assist in fighting COVID-19? This article might be an entry point for your inner brainstorm, although let’s not kid ourselves - if you are reading this newsletter you have probably thought about these things already.

    7) Publication of the Week - Lecture 23 Guest Lecture: Drones – CS287-FA19 Advanced Robotics at UC Berkeley (2020).
    YouTube
    INFO: If you are wondering how Skydio drones perform state estimation and navigation then this lecture by Hayk Martirosyan is for you. It’s fascinating that the depth estimation is running on their deep learning network at 1000 Hz. Another interesting insight is the camera calibration for thermal expansion (both for the camera location and the lens distortion). If you like this lecture then you might also like this behind the scenes video that was showcased in the issue #68 and is showing, among other things, automated camera calibration.

    1) Humble Book Bundle: Artificial Intelligence & Machine Learning by Morgan & Claypool.
    HumbleBundle
    INFO: “We’ve teamed up with Morgan & Claypool for our newest bundle! Get ebooks like Essentials of Game Theory, Natural Language Processing for Social Media, Creating Autonomous Vehicle Systems, and more” - as always with these book bundles if you decide to purchase it then you can choose to support this newsletter and make it better.

    Job Seekers

    In the issue #83 I’ve started this section to try to help out those looking for work in the times of pandemic. If you are currently looking for work then feel free to send me your details in the same format as you can see in the entries below. Please note that I will be able to list up to 5 profiles a week here.

    Name: Beau Everaert
    Location: Brugge, Belgium
    Skills: C/C++, ROS, Python, Gazebo, Linux, GIT, XACRO, Inventor, PLC Programming, Electric Drives, Microcontrollers, Assembly, JavaScript
    Profile: Graduated as a master in electromechanical engineering at the University of Ghent, I have experience as an automation consultant for a large enterprise and I am fluent in German, English and Dutch. I’m very eager to contribute, I learn quickly so wherever my skill set could be of use, I’d be happy to help!
    Social Profiles: LinkedIn
    Email: beau.everaert@gmail.com

    Announcements

    1) RobotUnion Marketplace.
    RobotUnion
    INFO: RobotUnion Marketplace is the showcasing platform for all the potential robotics stakeholders, which aims to become a single point of entry & to know-how services of technical and non-technical nature. All the RobotUnion Community users are invited to share their products and services in order to get visibility and create new business opportunities. A marketplace is also a tool that is used by RobotUnion scaleups for external services which they would like to use during the acceleration program.

    2) KUKA Innovation Award 2021.
    Kuka
    INFO: Apply now with an innovative concept to the “Artificial Intelligence Challenge”! You will have the opportunity to implement your concept with a KUKA robot provided free of charge and you will receive hardware training and coaching from KUKA experts throughout the competition. At the finale, which will take place in April 2021, you will have the opportunity to present your application to trade visitors from industry and research, the press, investors and interested visitors and win €20,000 in the end.

  • Weekly Robotics Q1 2020 report

    Here is some data for those of you interested in how the newsletter was doing in the last 3 months!

    Email Subscribers

    The growth rate of e-mail subscribers had slowed down quite a bit compared to last year’s: we’ve started the year with 1357 subscribers and we’ve finished the quarter with 1494 people on board. My spreadsheet says it’s 10% growth.

    Mailchimp growth

    Website Traffic

    Website visits

    The reported web traffic was down by about 27.5%. One probable cause of this is that the newsletter stopped appearing in Google Discovery.

    Time Spent on the Newsletter

    I’ve spent in total over 71 hours on the newsletter this quarter. This is not too far from the previous quarter’s 76:32 hours.

    Time spent on Weekly Robotics

    Highlights of the Quarter

    What I was really hoping to achieve this first quarter was to record another podcast episode. I had a pretty good line up that was crushed by the whole COVID-19 situation. The virus actually brought a tiny bit of freshness to the newsletter - as you might have seen I’ve started a Job Seekers section where roboticists looking for work can share their profile with the rest of the community. I’m planning to keep this section on until everything calms down - then I’ll most likely switch back to the old careers section.

    I’m always juggling with some ideas for this newsletter. Right now what’s itching me the most is the design - I’m really really tempted to start a redesign process of both the website and the e-mails. I’ll keep you posted on these developments!

    In other news: do you know these Humble Bundle ref links that I keep posting every now and then? Last quarter you donated $73 to WeeklyRobotics through these affiliate links. That is more than the whole 2019! Many thanks for your support and I want to assure you that every cent will be used on improving the newsletter.

    Stay safe out there! Mat

  • Weekly Robotics #85

    Issue 85

    There is a high chance you are in isolation/lockdown right now. If you feel talking about robotics can help lift your spirits then feel free to send me a message. The most clicked link last week was the Open Source Turtle Robot instuctable with 16.5% opens.

    1) Robots Learning to Move Like Animals.
    BAIR
    INFO: Researchers at Berkeley Artificial Intelligence Researcher (BAIR) and Google Research had developed a framework that enables legged robots to learn agile skills by imitating animals.

    2) Building My Own 50Ah LiFePO4 Lithium Battery Pack.
    The Life of Kenneth
    INFO: A very nicely written blog article on how to build a battery pack out of 4 LiFePo4 cells and putting them together with a charger in an ammo box.

    3) Results 2020 User Survey.
    ROS Discourse
    INFO: Katherine from Open Source Robotics Foundation shared the results of ROS 2020 user survey. Since I’m actively working on ROS and writing tutorials from time to time I found lots of insightful data in the survey. Hope you will too!

    4) Augean Robotics: Mechanizing Food Production From Farm To Table.
    Robot Rabbi
    INFO: “What on earth is a tractor doing at CES?” - this post by Oliver Mitchell gives a neat overview on robotics in agriculture in the US.

    5) I Am Mesmerized By Our New Robotic Vacuum.
    dev.to
    INFO: A very enjoyable read on a software developer being fascinated by his new low cost robotic vacuum cleaner.

    6) Ford Autonomous Vehicle Dataset.
    Ford
    INFO: Ford has published an autonomous vehicle dataset (for non-commercial use only). If you like laser scanners then you will like this one - each vehicle was equipped with 4 HDL-32E LiDARs. On top of that the dataset contains information IMU, GPS and 7 cameras. The data is provided as ROS bags.

    7) Design and Optimal Control of a Tiltrotor Micro Aerial Vehicle for Efficient Omnidirectional Flight.
    YouTube (aslteam)
    INFO: Normally this video together with this publication would make the “publication of the week”, however because I’m very short on time I didn’t have enough time to fully cover it. From the video and the paper you will learn about the omnidirectional multirotor designed by Researchers at Autonomous System Lab at ETH Zurich.

    1) ArduSimple RTK - ROS Integration.
    msadowski.github.io
    INFO: I got around to testing my u-blox F9P based ArduSimple RTK modules with ROS. In the above blog post I show a minimal integration with ROS that allows to visualize the data without any other hardware then the RTK units.

    Job Seekers

    In the issue #83 I’ve started this section to try to help out those looking for work in the times of pandemic. If you are currently looking for work then feel free to send me your details in the same format as you can see in the entries below. Please note that I will be able to list up to 5 profiles a week here.

    Name: Sagar Eligar
    Location: Boulder, CO, USA. Willing to relocate
    Skills: Python, Linux, MATLAB, C, ROS, GIT, Coppleia Robotics, Control Systems, and Embedded systems
    Profile: I’m a Robotics graduate student working on Mobile Robots, self-balancing robots, and control of robots. Previously I worked on an underwater vehicle AUV
    Social Profiles: LinkedIn
    Email: sagar.eligar@colorado.edu