Bimba: Sensor Embedded Balls for Creative Sound Generation
Pol Pla and Patricia Maes
Fluid Interfaces Group
MIT Media Lab
Cambridge, MA 02139 USA
{pol,pattie}@media.mit.edu
Abstract
Bimba is an exploration into designing playful interfaces
which use embedded sensors to sonify object interactions.
The system wirelessly collects data from sensors
integrated into foam balls and generates a sound composition
extracting the most relevant features of movement.
We have designed and built a collection of sensor
units, implemented a protocol to support a wireless real-time
data network infrastructure, and proposed a sonification
mapping for the generated output. The implementation
of Bimba informs design possibilities for
creative and expressive interfaces, with particular focus
on those generating sound.
 Introduction
Physicality has shown potential in facilitating the design of
complex interactive systems. Fitzmaurice, Ishii and Buxton
introduced the concept of graspable interfaces (1995) –
physical objects that interface with computers for more
meaningful human-computer interactions. Their work emphasizes
the central role that physicality plays in the design
of effective interactions. Further work (Ishii et al., 1997;
Maynes-Aminzade, 2003) has investigated how users can
leverage their acquired intuition about physical laws to
better understand tangible systems.
In our design, we seek natural and meaningful relationships
between physical properties and sounds. While there
is not a definitive way to sonify physicality, we are proposing
one approach with the aim of generating conversation
around the potential of such environments.
The intersection between tangible interaction and sound
generation has been explored before. Systems like the ReacTable
(Jordà et al., 2007) are examples of physical manipulation
of sound.
In the same vein, there exists a large collection of projects
that embed sensors into instruments (Young, 2002; Overholt,
2005) or into wearables (Marrin and Picard, 1998;
Paradiso et al., 2000) for the control of musical output or
even enhancing stage performances.
The system we propose builds on concepts and mappings
that we observed in these projects, but it is, at the same
time, distinct in two main aspects. Firstly, our system proposes
a duality between control and randomization, which
makes it different from tangible musical interfaces that
attempt to maximize control over the output. Secondly, it is
designed to be embedded in everyday objects – as opposed
to instruments, the performer herself, or her clothing.
There are two projects in particular that have deeply informed
our work: Squeezables (Weinberg and Gan 2001)
and Squidballs (Bregler et al., 2005). Squeezables are a
collection of soft instruments that generate sound in an
intimate way. Squidballs, on the other hand, involve a large
audience in massive collaborative play. We wanted to create
a system that works in both contexts, which is why we
implemented Bimba to suit both intimate environments as
well as shared spaces.
Expressiveness is an important factor in musical performance;
it helps the artist to better communicate her music
to the audience. At the same time, expressivity helps the
audience to better appreciate the performance – establishing
a connection between the performer’s actions and the
resulting sounds. We borrowed inspiration from projects
that try to emphasize the emotional aspect of a performance
when generating/affecting musical outcome such as
(Marrin and Picard 1998). We believe that the physicality
present in out system, combined with intuitive mappings,
can help convey emotions and contribute to an interesting
and informative visual – as well as auditory – experience.
As reflected in (André et al, 2009), serendipity can have a
great impact on the creative process. We already commented
on the affordances of object-based musical systems, but
we also believe that such systems offer interesting possibilities
for serendipitous exploration and sound randomization.
Physical laws can help us design more accessible systems,
but at the same time enable users to experiment with
uncontrolled environments. In doing so, performers may
even new inspiration in the surprising output that they generate.
For example, a performer might achieve unique results
by dropping one of these balls down a set of stairs or
juggling with a trio of them.
Our investigation, Bimba – Catalan slang for ball –, builds
upon research addressing the four characteristics described
Proceedings Proceedings of of the the Second Second International International Conference Conference on on Computational Computational Creativity Creativity 117
above: physicality, scale, expressiveness and serendipity.
Bimba is a system for sonification of everyday objects with
embedded sensors.
Technical Aspects of the System
Bimba consists of a set of small wireless powerautonomous
electronic circuits. The current configuration
of each circuit includes a microprocessor, a battery, a wireless
communication chip and three inputs – namely, a
three-axis accelerometer, an air pressure and temperature
sensor, and a piezoelectric sensor.
The goal of this design was to create an integrated system
that accommodates different physical objects collecting,
processing and broadcasting a diversity of real-time sensor
data. For the exploration discussed here, we incorporated
the circuits into foam balls roughly 15cm in diameter
Figure 1. One of the balls of the Bimba system and a prototype of
the sensor board.
We developed a communication protocol to process multiple
packets of data coming from sensors distributed across
different objects. Having built each of the boards with an
integrated wireless communication module, we created a
star topology network. That is, all communications are
centralized in a controller node. This controller node then
generates sound with the information received. This kind
of network enables us not only to generate music from
each individual sensor, but also to establish sound relationships
between all of the devices connected.
Physicality: Rules, Object Properties and
Mappings
In a system such as the one we are describing, it is important
to have clear constraints which guide the creative
process. The laws of physics, which are universally applicable
to the objects that surround us, provide common
ground for creation. The specific characteristics of each
object add the necessary variability to the system; depending
on its shape, weight, composition, and size, an object
interacts with the environment differently.
To achieve an appropriate balance between variability and
predictability, it is necessary to relate physical rules to
compositional rules in a meaningful way. For this reason,
mappings are an essential component of a successful system.
The system discussed here senses three different parameters:
acceleration, altitude and impact. By no means
are these the only interesting parameters to consider; for
future iterations of the system, we will design a customizable
circuit that enables addition and replacement of the
sensors included.
For each of the sensors used in the current version of the
system, we propose a set of mappings. These are described
in the following sections.
Acceleration. The system reads the aggregated acceleration
of the balls. This parameter is strongly related to the
sense of motion and momentum. Because movement implies
continuous variability, this parameter is especially
suitable for controlling the base frequency of the background
tone.
Altitude. The relative height is particularly well suited to
representing the primary sound’s pitch because it is intuitively
associated with a range from low to high.
Impact. Bouncing motion is clearly associated with percussion.
In our design, the system triggers a MIDI drum kit
sound each time a ball senses contact with another object
or surface.
Design Criteria and Goals
The system design was intended to explore four concepts
tied to our research interests: relating serendipity and inspiration,
supporting different creative styles, building immersive
environments, and enabling collaboration. We believe
that these four pillars lay the foundation for creative flow.
Serendipitous approach to creativity. As we alluded to
in the introduction, designing for serendipity is as important
as designing for controlled environments. By taking
advantage of environments familiar to the user, Bimba
supports playful exploration in a relaxed and intuitive way.
Ideally, this would minimize distraction so that the creative
experience may also be an immersive one.
Supporting different styles of use. Designing tools for
creation means providing resources in an unencumbered
manner so that users can challenge conventions of the system
in new and interesting ways. Research has shown the
importance of appealing to different creative styles to push
the boundaries of creativity (Resnick and Silverman 2005).
For this reason, Bimba aims to support a broad range of
approaches to sound generation.
Design that impacts the ecology. Successful designs affect
the ecologies that they inhabit. Since our system is
deeply connected to the environment – by means of the
Proceedings Proceedings of of the the Second Second International International Conference Conference on on Computational Computational Creativity Creativity 118
laws of physics – we expect users to modify spaces to
stage their own “worlds of music”.
From intimate use to massive collaboration. Bimba is
designed to support the full range between small-scale,
intimate interactions and massive collaborations. As previously
described, the system can be easily used for personal
exploration of sound landscapes. However, because the
proposed design is distributed, there are also many instances
in which it can be used collaboratively.
Discussion and Future Work
We talked about the importance of mappings for performers
to connect with an interface. Enabling users to create or
modify their own mappings may strengthen the bond between
the performer and the sound object.
The current system supports input from multiple objects,
but it does not yet propose mappings reliant upon inputs
distributed across a collection of objects. Improving these
distributed mappings may encourage higher levels of collaboration.

Additionally, if Bimba had an awareness of the material
composition of the objects, then the mappings could reflect
this. Zoran’s Chameleon Guitar (2009) is an inspiring example
of using resonance filters to represent different material
arrangements.
Another future line of research lies in creating a set of
modular sensor boards. Such a modular system would enable
performers to capture the most important aspects of
each object. For example, the physical characteristics of a
bouncing ball can be well represented by sensing motion,
altitude and impact – but this may not be the case when the
interface object is instead a glass bottle.
Acknowledgements
We would like to thank Joe Paradiso and Laurel Pardue for
providing a great collection of inspirational materials as
well as the technical guidance that enabled the production
of the work reflected in this document.
We would also like to acknowledge contribution of Emily
Lovell in editing and organizing the text in this document.
Conclusion
Bimba serves as a preliminary investigation into turning
everyday objects into musical interfaces through the use of
embedded sensing.
This paper reflects on the benefits of capitalizing on object
affordances to creative sound generation. By grounding our
exploration in four main design considerations – serendipitous
approach to creativity, supporting different styles of
use, designing for impacting the ecology, and cross-social
scenario – we hope to stimulate future dialogue around
them.
<references_biblio/>
References
André, M., Schraefel, M., Teevan, J., and Dumais, J. Designing
for (un)serendipity. In Proceedings of the 7th ACM
Creativity and Cognition Conference. 2009.
Bregler, C., Castiglia, C., DeVincezo, J., DuBois, R.L.,
Feeley, K., Igoe, T., Meyer, J., Naimark, M., Postelnicu,
A., and Rabinovich, M. Squidball: An Experiment in
Large-Scale Motion Capture and Game Design. In Intelligent
Technologies for Interactive Entertainment (Lecture
notes in computer science, 2005, Vol. 3814/2005, 23-33).
2005.
Fitzmaurice, G., Ishii, H., and Buxton, W. Bricks: Laying
the Foundations for Graspable User Interfaces. In CHI’95
Proceedings of the SIGCHI conference on Human factors
in computer systems. 1995.
Ishii, H., and Ullmer, B. Tangible Bits: Towards Seamless
Interfaces between People, Bits and Atoms. In CHI’97
Proceedings of the SIGCHI conference on Human factors
in computer systems. 1997.
Jordà, S., Geiger, G., Alonso, M., and Kaltenbrunner, M.
The reacTable: Exploring the Synergy between Live Music
Performance and Tabletop Tangible Interfaces. In TEI’07
Proceedings of the 1st international conference on Tangible
and embedded interaction. 2007.
Marrin, T., and Picard, R. The “Conductor’s Jacket”: A
Device For Recording Expressive Musical Gestures. In
Proceedings of the International Computer Music Conference.
1998.
Maynes-Aminzade, D. The actuated workbench: 2D actuation
in tabletop tangible interaction. M.S. Thesis, MIT Media
Lab, Cambridge, MA, 2003.
Overholt, D. The Overtone Violin. In NIME '05 Proceedings
of the 2005 conference on New interfaces for musical
expression. 2005.
Paradiso, J., Hsiao, K., Benbasat, A.Y., and Teegarden, Z.
Design and implementation of expressive footwear. In
IBM Systems Journal (511 – 529). 2000.
Resnick, M., and Silverman, B. Some Reflections on Designing
Construction Kits for Kids. In Proceedings of Interaction
Design and Children conference. 2005.
Weinberg, G., and Gan, S. The Squeezables: Toward an
Expressive and Interdependent Multi-player Musical Instrument.
In Computer Music Journal (MIT Libraries).
2001.
Young, D. The Hyperbow: A Precision Violin Interface. In
International Computer Music Conference. 2002.
Zoran, A. Chameleon Guitar : a physical heart in a digital
instrument. M.S. Thesis, MIT Media Lab, Cambridge, MA,
2009.
Proceedings Proceedings of of the the Second Second International International Conference Conference on on Computational Computational Creativity Creativity 119