LiveObjects: Leveraging Theatricality for an Expressive ... · LiveObjects: Leveraging Theatricality for an Expressive Internet of Things Cesar Torres University of California Berkeley

LiveObjects: Leveraging Theatricalityfor an Expressive Internet of Things

Cesar TorresUniversity of California BerkeleyBerkeley, CA 94720, [email protected]

Jasper O’LearyEric PaulosUniversity of California BerkeleyBerkeley, CA 94720, [email protected]@berkeley.edu

BA C

Figure 1: a) LightWire, an illuminated, spinning optic fiber thatforms light bodies from social data, b) GorgeBox, an object thatinflates based on the accumulation of data, c) LiveFans, two IoTcontrollable objects that actuate relational data.

Permission to make digital or hard copies of part or all of this work for personal orclassroom use is granted without fee provided that copies are not made or distributedfor profit or commercial advantage and that copies bear this notice and the full citationon the first page. Copyrights for third-party components of this work must be honored.For all other uses, contact the Owner/Author.Copyright is held by the owner/author(s).DIS’16 Companion, June 04-08, 2016, Brisbane, QLD, Australia.ACM 978-1-4503-4315-2/16/06.http://dx.doi.org/10.1145/2908805.2908807

AbstractLiveObjects approaches expressive object design from thelens of art theory utilizing theatricality, or the perception ofan object having presence, to create a series of IoT dataobjects. This approach actuates materials and everydayobjects to express social data and provokes new interac-tions between objects, viewers, and space. We showcasehow such objects can form unique expressive personalities,expose relational data, and perturb environments to forminformation spaces.

Author KeywordsAmbient devices; data objects; data sculpture; IoT

ACM Classification KeywordsH.5.m [Information interfaces and presentation (e.g., HCI)]:Miscellaneous

Introduction and BackgroundAs the Internet of Things (IoT) or the vision of ubiquitousconnected everyday devices develops, we see opportunitiesto expand the design landscape of devices to go beyondneutral, atomic, functional, and obedient “things”. Such de-vices process functional bits of information and mirror thedata they receive in solely the space they occupy; inter-actions are often limited to quick punctuated glances withwearable devices, handhelds, and ambient screens.

Demonstration DIS 2016 Companion, June 4–8, 2016, Brisbane, Australia

5

Early work in ambient displays such as Jeremijenko’s LiveWire [4] and Ishii’s ambientRoom [3] showcase the value ofexpressing data in the cognitive background through ambi-ent actuation. Data sculptures, or data-based physical arti-facts, revealed that materializing data using the right formsand mappings can enhance a user’s understanding of dataand the social issues that underlie it [13]. Furthermore, asan alternative vision of “smart” is developing, a small sub-set of objects that co-inhabit our space are challenging theconstrained objecthood of devices. As simple as namingdevices, developing more intimate interactions have beenshown to increase pleasure, social inclusion, and personalvalue in devices such as the Roomba vacuum [10]. Recentwork such as AniThings [12] proposed animistic devicesthat have less imposing, yet unique personalities defined bytheir capabilities and limitations. Similarly, the Living Inter-faces utilize zoomorphism and anthropomorphism to createsocially-expressive robotics (a water-saving faucet[11], aneedy toaster [1], and a kiss-operated lock [8]).

Feature

Phrase CountSentiment (Polarity)a

Sentiment (Subjectivity)

CountVelocityAcceleration

Relative CountRelative VelocityRelative Acceleration

Table 1: Feature list from ourTwitter scraper.

aPython NLTK Sentiment Analysis

In contrast, LiveObjects approaches expressive object de-sign from the lens of art theory and creates a set of IoTdata objects that utilize theatricality, or the perception of anobject having presence, to provoke new interactions whereobjects are active agents in a space. Theatricality refers toa term coined to critique Minimalist art objects in the 1960s(c.f. [7, 5, 2]). The three-dimensionality of these objectsadded a “theatricality”, or an awareness of the object; suchawareness would disrupt the “absorption” of the object withthe viewer. For instance, a precarious one-ton assemblyof metal plates [9] might elicit careful movement from au-diences and exude a presence equivalent or greater thanother bodies in that space. This “presence” carries largeimplications to proxemics, or the manner in which proximitywith an object forms zones of interaction that play a largerole in audience engagement [6].

LIVE OBJECTSWe created a series of IoT objects that explore theatrical-ity through the use of material and space. All three ob-jects track twitter data streams and express this social datathrough activation of materials and actuation. We describeour data tool and the design decisions, construction, andinteractions with these objects.

Twitter as mediumWe constrained the scope of our data source to Twitter fora variety of reasons: a) it is rich in relevant social and sci-entific data (e.g. @USGSted tweets up-to-date earthquakealerts), b) the data is diverse in subjectivity, and c) it has ahighly accessible streaming API. As a brief overview, ourtool is able to access tweets by person or entity (@), tags(#), and mentions. A select set of tweet data streams arelogged into a database and the extracted features (Table 1)are exposed through a command-line interface. A PythonAPI interface communicates to objects via Bluetooth; eachobject uses a Bluetooth Serial Port Module (JY-MCU).

LightWireWe draw inspiration from Natalie Jereminjenko’s 1995 pieceLiveWire (also known as Dangling String), an evocativedata object consisting of a simple wire that vibrates as afunction of network traffic data and operates in the cogni-tive background. For example, a user might be uploadinga file, notice a significant delay, and discern from a rapidlyoscillating wire in her periphery that network usage is high.

LightWire is a similar IoT object, yet it oscillates an opti-cal fiber around an axis which creates light “bodies” frompersistence-of-vision (Figure 2). At one end of the opti-cal fiber (SGS2 5/64"), a DC motor is attached using a 3Dprinted motor arm; at the other end, the fiber is threadedthrough a support block and affixed to an enclosure con-taining an LED and battery, allowing the wire to free-spin.

Demonstration DIS 2016 Companion, June 4–8, 2016, Brisbane, Australia

6

By either varying motor speeed or altering the tension ofthe wire, we are able to create diverse waveforms.

This particular optic fiber consists of a solid plastic core witha thinner cladding (a material with a lower refractive index)which reflects light along its side. As the fiber bends fromthe oscillations, the amount of refracted light changes as afunction of the bend radius. The analog factors of the wire’sspin, such as its inertia and form, create imperfections thataffect the versimilitude of the data it receives.

optic fiber

DC Motor

LED

µC + Bluetooth

Figure 2: LightWire, an IoT objectof configurable scale that oscillatesan optic fiber back-and-forth.

Balloon

InterchangeableMaterial Frames

Chasis

INFLATED STATE

DEFLATED STATE

Figure 3: GorgeBox, asoft-actuated IoT object withinterchangeable material frames.

Although directly mapped to the data, from an observer’spoint-of-view the legibility of the data is ambiguous. For in-stance, mapping the earthquake data 1:1 to the dynamicrange of wire oscillations will cause the wire to accuratelydisplay the data, however each data point will be subject tothe last (e.g. a series of consequtive large earthquakeswould appear very different from an earthquake over asilent period). In this way, the object takes on an inertialbias that high-fidelity digital displays lack. From a theatrical-ity standpoint, the object achieves a notable presence fromits sheer scale. In our construction we use a 3 meter fiber,which needs to attach itself to a rafter of a room, embed-ding itself and subject to the architectural space. The objectinhabits that space, actively requiring a radius of space tooperate and creates light bodies which require a “personalbubble” roughly equivalent to human agents. The visceraldiverse forms it creates contributes to its liveliness, andwhile characterizable as an ambient device, it elicits periph-eral attention.

GorgeBoxGorgeBox pits two materials against each other: the struc-tural, seriousness of wood, and the playfulness of a rubberballoon. We construct a cuboid structure out of birch. Eachface is formed from an interchangeable material frames. Inour construction, we stretch different fabrics (an elastic ny-

lon, a polyester mesh, and matte tulle) over these frames.Lastly, a rubber balloon and pneumatic control mechanismis inserted in the structure. The balloon is inflatable usingan air pump where a digital control valve is used to maintainair pressure.

In our construction, we chose to implement a tweet “inbox”where we collected all mentions of a user over the courseof the day. This could be in the form of direct messaging orside channel chatter. We mapped these values to outputvolumes of air which would continue to accumulate andinflate the balloon, “gorging” on this gossip, unless the userresponded to these messages.

The GorgeBox makes use of the dynamic affordances ofthe balloon. As the balloon expands, it enters a differentstate of being; the tension of a balloon on the verge of ex-plosion is a unique experience that captures attention. Byconstraining the balloon inside the frame, this expansion(and tension) is exacerbated and dramatized, eliciting animmediate user interaction. The balloon does pop if unat-tended, which offers a unique method of achieving “live-ness” from the contrast of a exhausted object.

LiveFansA pair of commodity fans with four-level (HIGH, MED, LOW,OFF) speed control were repurposed with a set of relayscontrolled by an Arduino microcontroller (Figure 4).

In the current space of IoT devices, control is often limitedto the scope of a single object. As a pair, LiveFans explorehow multi-actor systems can be leveraged to display rela-tional data and collaboratively actuate an enviroment. Themovement of air extends the scope of the data to the envi-ronment where this actuation can be felt by an audience inthe vicinity of the fans.

Demonstration DIS 2016 Companion, June 4–8, 2016, Brisbane, Australia

7

As such, the data can form a dialectic; in the case of thefans, we control respective intensities using oppositionaldata. To showcase this, we harvested data from fifty of thetop US politics Twitter accounts and searched for the rela-tive acceleration of tweets mentioning the top two politicalcandidates during an election season.Each fan is assigneda political affiliation and spins according to live tweet data.As such, a new subjectivity is introduced to these objects.This configuration forms a political “microclimate” that canbe experienced by several users as an engaging, embodiedinteraction.

@HillaryClinton @BernieSanders

Figure 4: LiveFans, repurposedcommodity fans turned into IoTdevice and sent oppositional datafeeds.

ConclusionIn this work, we demonstrate the design of objects with the-atricality, or presence, by juxtaposing actuation, space, andmaterial. Through this exploration we show how objectscan exhibit a unique bias and personality, that the spaceof interaction can far exceed the bounds of the object tocreate information spaces, and that how materials can beused to dramatize interactions. At DIS, we invite attendeesto experience how these theatrical IoT objects can engageaudiences with data.

AcknowledgementsWe thank Connor Freeman, Emily Chou, and Neil Kumarfor their aid with fabricating live objects. This research wassupported by NSF Grant No. IIS-1451465.

References[1] Eva Burneleit, Fabian Hemmert, and Reto Wettach.

2009. Living Interfaces: The Impatient Toaster. InProc. of TEI’09. ACM, New York, NY, USA, 21–22.

[2] Michael Fried. 1967. Art and objecthood. Art andObjecthood (Chicago, 1998) 153 (1967), 165.

[3] Hiroshi Ishii, Craig Wisneski, Scott Brave, AndrewDahley, Matt Gorbet, Brygg Ullmer, and Paul Yarin.

1998. ambientROOM: Integrating Ambient Media withArchitectural Space. In Proc. of CHI’98. ACM, NewYork, NY, USA, 173–174.

[4] Natalie Jeremijenko. 1995. Live Wire (DanglingString). (1995).

[5] Donald Judd. 1965. Specific objects. Arts Yearbook 8(1965), 74–82.

[6] Daniel Michelis and Jörg Müller. 2011. The audiencefunnel: Observations of gesture based interaction withmultiple large displays in a city center. Intl. Journal ofHCI 27, 6 (2011), 562–579.

[7] Morris Robert. 1966. Notes on Sculpture. Art inTheory 1900–1990: An Anthology of Changing Ideas(1966), 813–822.

[8] Miriam Roy, Fabian Hemmert, and Reto Wettach.2009. Living Interfaces: The Intimate Door Lock. InProc. of TEI’09. ACM, New York, NY, USA, 45–46.

[9] Richard Serra. 1969. One Ton Prop (House of Cards).(1969).

[10] Ja-Young Sung, Lan Guo, Rebecca E. Grinter, andHenrik I. Christensen. 2007. "My Roomba is Rambo":Intimate Home Appliances. In Proc. of UbiComp ’07.Springer-Verlag, Berlin, Heidelberg, 145–162.

[11] Jonas Togler, Fabian Hemmert, and Reto Wettach.2009. Living Interfaces: The Thrifty Faucet. In Proc. ofTEI’09. ACM, New York, NY, USA, 43–44.

[12] Philip van Allen, Joshua McVeigh-Schultz, BrooklynBrown, Hye Mi Kim, and Daniel Lara. 2013. AniThings:Animism and Heterogeneous Multiplicity. In CHI ’13Extended Abstracts. ACM, New York, NY, USA, 2247–2256.

[13] Jack Zhao and Andrew Vande Moere. 2008. Embod-iment in data sculpture: a model of the physical visu-alization of information. In Proc. DIMEA. ACM, 343–350.

Demonstration DIS 2016 Companion, June 4–8, 2016, Brisbane, Australia

8