top of page

DC Protest Data INfrastructure

Protest Infra.jpg

How might the digital footprints from our social media be used as a source of bottom-up, urban planning tool for cities?


Rally Infrastructure :: Social Media Data in Urban Design

Harvard GSD Research Studio :: Spring 2019

Instructors :: Grace La, James Dallman

     Part of the winning portfolio submission for the GSD Digital Design Prize 2021

This project looks at data-driven strategies for site analysis and urban analysis, mining social media posts as a data source to drive siting for a series of political rally infrastructure points in Washington, DC.

Problem Space :: Urban Experience of Rallies after Instagram

Current rally infrastructure in dc consists largely of massive, empty green spaces that, when not in use as gathering spaces for rally events, are designed more to frame views between monuments than for actual occupancy.

This type of infrastructure was well-suited to rallies in the 60’s 70’s and 80’s, where one’s experience of the event centered around occupying the same physical location as thousands of other people, andwhere the media presence of the event depended on singular,  large shots of massive crowds in a single location.

However, these types of spaces don’t necessarily resonate with today’s experience of political demonstrations and rallies. In “smart cities,” Antoine Picón describes a certain type of “bottom up, reactionary, spontaneous” character to contemporary urban events fueled by new communication technologies; Contemporary political rallies in DC seem to strongly reflect these trends. The way people relate to the event through social media and photos has dramatically changed, and the way the event is experienced has changed in a way that no longer relies on a single, symmetrical space framing a stage view.

The intent of this project is to use this phenomena of the new, digitally augmented rally experience to explore architectural ideas of collective urban form and a more dispersed institutional identity that better supports the changing character of these events.The project proposes a new institutional type (institute for civic engagement) that both infrastructurally and spatially supports this new, more spontaneous, dispersed, weak-figure urban experience of the rally events, and creates a network of small museums between prominent contemporary and historic rally sites.


1963: Large crowd gathered in front of Lincoln Memorial during Martin Luther King Jr's "I Have a Dream" speech.

image-asset (1).jpeg

2008: Women's march protesters document their experience of the rally in many parts of Washington DC at once, through Instagram :: Stage-like spaces and fields for crowds to gather in a single place are less important to 21st-century protesters.

Data-Driven Siting

Social Media Data Gifs.gif

To find the optimal siting locations, this project mined a series of social media posts on different topics, such as the feminist Women’s March, the gun-control-supporting March for Our Lives, or the conservative rallying Unite the Right. We designed a set of digital tools to map out the location of these social media posts, learning when, where, and how these more dispersed, bottom-up rallies occupied the city (see image, above).

We mapped out the digital footprint of these rallies to understand the difference between pre- and post-digital urban experience of political rallies (right). One can clearly see the fragmentation and dissolution of contemporary rally spaces relative to historic ones.

A set of other data mappings, such as car-traffic routes and protest-related road closure areas, complemented our social media dataset — creating pedestrian crossing points at these intersections could alleviate the problem of road closure on DC’s busiest streets. These


Urban footprint of rally events plotted along a timeline. While rallies occupy the same general area of the city, their footprint has dramatically spread out as social media becomes more popular. Click to flip between a timeline and mapped image.

moves related to infrastructure and street crossing try to reduce the extremely complex permitting process required to host a rally, and by extension encourage the more spontaneous and informal characteristics of these events that new communication technologies have already started to create on their own.

traffic data.jpg
street closure heatmap.jpg

Finally, we considered visibility across the DC Mall's long open spaces and onto its most important historic rally locations (see image below). The siting locations synthesize these and our other data sources, so that from each pavilion’s crossing point, there is an unobstructed view to the next pavilion, allowing visitors to easily follow a route between them. 

Siting Locations.jpg

A Museum Dispersed Throughout DC


The project is tested in a proposal for the “institute of civic engagement,” a program that, on days with large political rally events, formalizes and makes permanent the currently transient, temporary, throw-away infrastructure that is quickly put up and taken down for rallies, and on days of “normal” operation creates a continuous museum experience which is educational about the American tradition of civic engagement through rallies in DC. 

Functionally, the objects try to accommodate the rally events themselves but also the massive amounts of infrastructure that is temporarily shipped in and then dismantled every time an event happens; For example, large numbers of restrooms, AV equipment, or news reporting equipment, all of which are quickly moved in and disassembled every time even a small event happens.


Different museum (left) and rally (right) functions our Institute for Civic Engagement takes on.

While many contemporary museum projects in DC are formally quite similar to the traditional neoclassical architecture of dc in their bilateral symmetry and tripartite division, The institute’s formal character is inspired more by late-deconstructivists’ and early-parametricists’ experiments in more a dispersed notions of formal or institutional identity and urban form.  The project pulls from archetypical references such as Tschumi’s Parc de la Vilette, which test ideas of geometric resonance between separate follies, or ZHA’s Innsbruck station, which carefully controls the genotype/phenotype relationships between separate stations to communicate a common formal identity between station entrances with sometimes dramatically different functional requirements.

While the total museum area is similar to that of  the recently completed NAAMHC, it is split between these physically disconnected sites. The experience of moving between “levels” of curatorially disconnected exhibition space in NAAMHC is replaced with an experience of moving between these road-crossing objects as one traverses this area of DC. 


Our Museum of Civic Engagement Massings (left) hold a similar program area to the recently-completed NAAMHC museum (right), but dispersed across the city.


A Parametric Network

The parametric system for generating each crossing had three main goals: To give them a common visual identity, that makes them recognizable as a series of related institutions; To guide visitors between each pavilion as one moves through the national mall area; and to achieve structurally-feasible spans, sometimes with significant cantilevers over major roadways.


Large-scale geometric alignments between these objects reinforce the experience of these relationships, sometimes between locations that are more than a mile apart yet are still visible because of DC’s flatness and it’s strong major axes, create the impression of a continuous building experience between them. The raised vantage points the crossings create allow for views between sites, photo or news reporting points, and a sectional doubling of the density of people that can participate at each site (see image, right).

Structurally, each pavilion is hung off of a square tube truss that spans each crossing. The tube-structure at each crossing is simulated computationally to “balance” on each crossing (see image, right). The structure’s massing is manipulated to see which cantilevers will counterbalance the structure, allowing each crossing to achieve larger overhangs and take on more unusual formal characteristics at each crossing.

views between sites.jpg
center of mass.jpg

Objects frame views between sites, so visitors can easily navigate between them.

The massing was structurally simulated in a computer physics engine to find optimal balancing points for our cantilevered volumes.

The Institute of Civic Engagement

Three sites were developed into full architectural concept solutions, testing ways the tube structures would lend themselves both to novel museum experiences and interesting functionality during rallies. These pavilions explore new types of collective urban form that vibrate between legibility as single, folly-like objects and continuous urban experience.

Site 01: "Stagefront" Protest Infrastructure

site 1.jpg

Inspired by contemporary rally’s emphasis on framed stage views and digital telepresence, the first location frames the capitol building during a rally, and creates a large outdoor amphitheater that projects views of other rally locations onto the building. In its museum function, the structure showcases mixed-media and digital artwork resonating with the nearby National Gallery of Art buildings

site 1 ref.jpg

Site 03: "Crossing" Protest Infrastructure

Site 3.jpg

Drawing from the value of raised vantage points for rally news coverage and stage views, site 3 creates an elevated park that extends John Marshall Park into Freedom Plaza, and gives protesters a prime space for poster and banner display over the high-traffic 14th st. 

site 3 ref.jpg

Site 05: "Roofscape" Protest Infrastructure

Site 05.jpg

Inspired by historic rally locations along the national mall, site 5 pulls the landscape of the mall up onto its roof and over 14th st. In its museum function, this site reflects on the success of recent underground museum extensions around the mall, such as BIG’s Smithsonian extension, and creates a link to the subterranean gallery space of the adjacent NMAAHC.

site 5 ref.jpg

In summary, the project tests three design hypotheses: First, to demonstrates the potentials of mining bottom-up data sources such as social media posts as an urban design tool; Secondly, to leverage generative, parametric design tools to create common formal identity between physically disconnected objects; and Third, to test new forms of institutional identity that disperse through the city.

Click for Full-Sized Gallery:

bottom of page