Prototyping DIY smart robots with Arduino and Machine Learning

Ivan Iovine

The workshop aims to teach participants the use of the Arduino platform in conjunction with the Ml5.js Machine Learning framework.

Each participant will be given a DIY robotic arm made of recycled wood, to which an Arduino will be interfaced. Through serial (WebSerial) communication, the Arduino will communicate with a Javascript application and the Ml5.js framework. Participants will be explained and taught the basics of Machine Learning, exploring and experimenting firsthand with pre-trained Machine Learning models for body recognition (PoseNet model), hand recognition (Handpose model), face and facial emotion recognition (FaceApi), as well as real-time object tracking (YOLO). Through the use of these Open Source technologies, workshop participants will be able to learn the basics of Arduino and Ml5.js, experimenting in a hands-on manner and creating customized human-machine interactions based on Machine Learning models.


Ivan Iovine is an interaction designer and media artist currently based in Frankfurt am Main. His artistic research focuses on the field of physical computing, physical interaction, machine learning and robotics. His works have been exhibited at “Maker Faire Europe” in Rome (2016), “JSNation Conference” in Amsterdam (2019), “Lab30” in Augsburg (2019) “Piksel Festival” in Bergen (2020) and “Die Digitale” in Düsseldorf (2021). In 2021 his work Theodore was published in the official journal proceedings of the scientific symposium “Art Machines 2: International Symposium on Machine Learning and Art”. The artwork was then exhibited in the same year at the Run Run Shaw Creative Media Centre in Hong Kong. He is head of the robotics lab at HfG Offenbach and teaches Physical Computing in the university’s art faculty.

Open Wave-Receiver

Shortwave Collective

Building Open Wave-Receivers enables DIY communications reception, and allows anyone to freely listen to the broad spectrum of radio waves around us. All you need are a few easy-to-procure supplies and, if you want to try it, a neighborhood fence or other receptive antenna proxy.

Why a fence? Antennas are necessary for radios to receive signals, and many things can be antennas. Fences can make great, and very long, antennas! Other materials can work well too; even a tent peg can become a useful part of a radio. Open Wave-Receivers allow us to explore the relationship between different combinations of materials, antennas, and radio waves, creating a new technology literacy, a new medium for artistic expression, and a new way to explore the airwaves in our communities.

We have found making Open Wave-Receivers to be a fun adventure. The ability to use simple scraps to create variety and personalization in each radio makes this a great maker project for anyone wanting to play with radio.


Shortwave Collective is an international, feminist artist group established in May 2020, interested in the creative use of radio. We meet regularly to discuss feminist approaches to amatuer radio and the radio spectrum as artistic material, sharing resources, considering DIY approaches and inclusive structures. Members include Alyssa Moxley, Georgia Muenster, Brigitte Hart, Kate Donovan, Maria Papadomanolaki, Sally Applin, Lisa Hall, Sasha Engelmann, Franchesca Casauay, and Hannah Kemp-Welch

Live collaborative radio with Mezcal

August Black

Mezcal is a web app for collaborative sound and live transmission that I have been prototyping and building in collaboration with https://wavefarm.org and multiple artists (such as Anna Friz https://nicelittlestatic.com/, Betsey Biggs https://www.betseybiggs.org/, and Peter Courtemanche http://absolutevalueofnoise.ca/?now).  In this 1 hour workshop, I give an overview of the software, its design intentions and practical implementations, and then split the group up into sections to create a live experimental radio session on-site. (note: this software is not YET free software, but lives in the web as a free service for free cultural institutions such as radio libre in Medellín, Colombia https://red.radiolibre.cc/ and Sound Camp in the UK https://soundtent.org/, among others)
https://august.black/mezcal/


August Black is a hybrid practitioner of art, design and engineering. He makes experimental spatial and acoustic situations, often by building his own technological artifacts and instruments in hardware and software. His past work focused on live networked audio, mixing FM radio with user input through online software. His current interests span the fields of the philosophy of technology, software studies, techno-politics, peer-to-peer networking and AI/machine learning. In the past, he’s been a member of arts organizations such as the ORF Kunstradio and the Ars Electronica Futurelab, as well as a former member of the engineering team at Cycling ‘74, makers of Max/MSP. He has shown works at festivals and venues such as Ars Electronica Festival, Dutch Electronic Arts Festival, Wave Farm, Transmediale, Pixelache, LA Freewaves, Piksel Festival, Polar Circuit and the Tasmanian Museum of Art, among others.He earned a BFA at Syracuse University and was an NSF IGERT Fellow at UC Santa Barbara, where he completed an MS and PhD. He’s taught media and art classes at UC Santa Barbara, University of San Francisco and CU Boulder, where he serves as Assistant Professor of Critical Media Practices.

https://august.black

memoryMechanics

Sensors and Pd

Kris Kuldkepp

Lecture and workshop on sensors and motion capture in new music and
multimedia performance.

The performer’s body is classically considered secondary in presenting a
musical piece. After all, the composer’s name and the idea of self-contained
artwork are predominantly the reasons for a concert visit rather than particular
performer(s) and their performing modes. Thus, for a classically trained
musician, it presents a conflict of being essentially a practical tool in the
service of a composer rather than a creative agent. However, the research in various music performance studies has resulted in a diverse quest for the importance of musical gestures and performers’ physical movements in transmitting the meaning. Do some movements or gestures of a
performer make the music meaningful? How could one classify musical gestures? How is acousmatic music perceived? How to understand electronic music in which the sound production is decoupled from physical gesture? … These and more are the open questions that circle in musical gesture research.
The focus on the importance of musical gestures has influenced composers and performing musicians to create pieces and improvisations that experiment with motion capture and various sensors to create musical experiences. The lecture introduces the usage of sensors and motor capture systems in new music and multimedia and discusses the philosophical concepts motivating the
research.

The lecture is followed by a practical workshop for absolute beginners in Pure Data (Pd) and programming for movement sensors.It will introduce the first steps for processing the data and the necessary algorithms.

The participants should bring their personal computers and preinstall Pure Data Vanilla (https://puredata.info/downloads/pure-data) as well as external libraries “Cyclone” and “else”.

In order to install the externals, please start up Pure Data and use Pd’s own external download manager:
— go to the “Help” menu
— choose the “find externals” option under the Help menu
— search for “cyclone”.
— Then click on the link to download the “cyclone” to your computer and specify the search path. (By default it should be ~/Documents/Pd/externals).
— Do the same process with the “else” externals.

Now the external libraries are installed in your computer

In order to load the libraries at the startup:
— go to “Preferences” menu in Pd
— choose Startup”,
— then click “New”, type “cyclone” and hit OK.
— Do the same with “else” library.
— Next time you restart Pd, the libraries will be loaded and ready to use.

Also, please download this folder with sound files into your computer. We use these sound files during the workshop as examples.
https://drive.google.com/drive/folders/12znhQbAJnaXWWrMLFPtIS6CZ9h85QhPP

As sensors, we will use our smartphones, and participants should also preinstall an app PdParty (iOS) or
Sensor2OSC (Android) on their phones. A computer mouse can be used to
stimulate the data stream.
During the workshop, we will build examples in Pure Data that introduce the
first essential steps in creating music with sensors and what to do with the raw
data. No previous experience with Pure Data is required.


Kris Kuldkepp

Kris is Hamburg based free improviser performing on double bass, bass guitar, and live-electronics. She is a feminist performer and artist. Kris is currently completing doctoral studies at Hamburg University of Applied Science where she researches spatial sound, free improvised music, and posthumanism. She is an active soloist and ensemble performer and has participated in festivals such as StimmeX, Blurred Edges Festival, and Bruital Furore in Hamburg, LjudOLjud in Stockholm, Tallinn Music Week, St. Petersburg New Music Festival reMusic, Estonian Music Days, she has also been invited to collaborate in various constellations within Europe and throughout South-America.

Kris is also part of the free improvisation quartet ‘double bird’ that recently released an album ‘favourite galaxy’ and is active with the quartet EMN concentrating on performing graphic scores and performative compositions. Kris is a close collaborator of opera director Lisa Pottstock with whom she develops feminist performances focusing on finding new way of dealing with body, materiality, and sound.

Ewasteroid

Paul Granjon

The beauty and the ugliness of electronic waste fight it off in this workshop for curious people. Starting with a pile of electronic waste items such as printers, pc towers, DVD players the participants will build a spinning asteroid made of out of date components and found timber, mining the old machines for intricate and complex parts. The resulting temporary sculpture is both celebration of human engineering and sinister indicator of an extractivist civilisation gone in overdrive.

Expect improvisation, technological creativity, freestyle wiring, collaboration and low-tech solutions. In line with Granjon’s current methods, the machine will work off-grid, be made of 90% recycled or found components and use open source technology controllers (Arduino).

The Ewasteroid belongs to Granjon’s extensive practice of Wrekshops, participat. The events combine hands-on, fun making with grassroots conversations inspired by the material, its abundance and creative potential.

The participants do not need to have prior knowledge of electronics or programming, start age 7 (under 12 accompanied by an adult). The workshop can run for a few hours or a whole day or 2, with participants coming and going, or booking a slot. Max 7 participants at a time with 1 assistant.

The Ewasteroid can be exhibited as an installation after the workshop, before its parts return to the recycling plant.

The first Ewasteroid was tested during the Deershed Festival in the UK in July 2022.


Paul Granjon is interested in the co-evolution of humans and machines, imagining solutions for alternative futures and sharing his experience of creative technologies. He has been making robots and other machines for exhibitions and performances since 1996. Granjon’s work became known for a trademark combination of humour and serious questions, delivered with absurd machines that made use of recycled components. His Sexed Robots were exhibited in the Welsh Pavilion at the Venice Biennale in 2005. He performs and exhibits internationally, with recent commissions in Garage Museum Moscow and Azkuna Zentroa Bilbao. He regularly delivers Wrekshops, public events where participants are invited to take apart electronic waste and build temporary new machines from the bits they find. Granjon’s current work is driven by an ecologist and participatory agenda. He teaches Fine-Art in Cardiff School of Art and Design, UK and completed a practice-based PhD in robotic arts in 2022.