The Cartography of Invisible Lines

For two years I walked neighborhoods, sat in cafes, and recorded conversations with strangers about the invisible lines they cross every day: the street where you suddenly feel watched, the park bench that belongs to someone else, the block you avoid after dark. This project turns those ephemeral boundaries into tactile, detailed embroidered maps on translucent silk. Each map starts with a raw GPS track of the person's recounted route, overlaid with a grid of their spoken coordinates—'here I felt safe, here I quickened my step.' Using a custom algorithm I wrote in Processing, I convert the audio waveform into a thread path, then hand-embroider or machine-stitch the resulting line work onto silk panels. The finished pieces are backlit, so the thread casts shadows that mirror the original anxiety or comfort. The project is a one-person show: I am the fieldworker, coder, embroiderer, and curator. It deliberately excludes any collaborative team, because the intimacy of the interviews and the hand-work demands a single, consistent presence.

We have maps of streets and buildings, but no maps of the emotional and social boundaries that shape our behaviour. Traditional cartography flattens experience into neutral geometry. This project scratches the itch to make visible the invisible rules that govern public space—the lines we all feel but never draw.

• Collect 30+ in-depth interviews with residents of one diverse city (Chicago used as prototype)
• Develop a Processing script that converts audio amplitude peaks and GPS waypoints into embroidery thread coordinates
• Produce 10 finished silk maps (50x70 cm each) backlit in custom lightboxes
• Mount a solo exhibition in a non-gallery space (e.g. a library or community centre) to invite public reflection
• Generate an open-source guide for other solo artists to replicate the method

• Not a scientific or sociological study—no statistical analysis or IRB approval sought
• Not a digital interactive or app; the maps are physical and analog
• Not a collaborative or crowdsourced project; all interviews and embroidery done solo to preserve consistent voice
• Not about objective truth—each map is consciously subjective and poetic

• Field recording: Zoom H4n audio recorder, Garmin eTrex 10 GPS (worn on the interviewee's belt during walk)
• Data processing: Processing 4 (Java) for audio-to-coordinate conversion; Python for cleaning GPS logs; Audacity for audio cleanup
• Marking up: Printed grid templates on tracing paper; fabric markers for transfer
• Embroidery: Brother SE1900 embroidery/sewing machine + hand needles; DMC floss in 12 colours (all blacks and greys with one accent red)
• Substrate: 100% silk habotai (pre-stretched on wooden frames)
• Backlighting: LED edge-lit frames with diffusers, custom-built with Arduino-controlled dimming (to simulate time-of-day mood)

The project follows a three-phase method. Phase 1: Capture – Each participant wears a GPS logger and a lapel mic while walking a route they normally take. Afterwards, I conduct a 20-minute interview where we listen to the recording together; I ask them to identify moments of boundary-crossing (felt safety, unease, ownership). I log those timestamps. Phase 2: Translate – In Processing, I overlay the GPS path over a street map of the neighbourhood. The audio amplitude from key moments becomes a sine-wave pattern that displaces the map boundary outward (louder = more displacement) like a contour line. I then convert that displacement into a stitch file (DST format for the embroidery machine). Phase 3: Embroider & Display – I transfer the stitch file to the machine, embroider the base map outline in black, then hand-stitch the displacement lines using free-motion embroidery to keep a human touch. The silk is stretched over a lightbox; dimming simulates the time of day of the original walk. Multiple maps can be stacked to show layered boundaries.

• Interviewee discomfort: Sharing sensitive routes may trigger anxiety. Mitigation: full anonymity, no real names, and skip-any-question protocol.
• Technical failure: Embroidery machine jams on silk; GPS drift in urban canyons. Mitigation: test on cheap fabric first, use GPS averaging.
• Artistic failure: The maps may look like chaotic thread tangles, not legible. Mitigation: early prototypes and peer critique with other textile artists.
• Time drain: Solo embroidery is slow; 10 maps could take 9 months. Mitigation: batch tasks, use machine for base, hand only for accents.

• Can the algorithm distinguish between 'good' tension (alert) and 'bad' tension (fear) from audio alone? Should I add heart-rate data?
• Do I need to reveal the exact street names on the maps, or distort them to protect privacy? Distortion might break the 'real map' feeling.
• Should the lightbox dimming be automated via a script that replays the original audio? That would require syncing and installation complexity.
• Is a gallery show the right endpoint, or should the maps live in the neighbourhoods they depict (e.g. installed in bus shelters)?

I finished three prototypes and five interviews before my funding ran out and I accepted a full-time teaching job. The project is still on my hard drive—the code, the silk, the recordings. Maybe one summer I'll pick it up again, but for now it's a beautiful what-if that taught me how to listen and sew at the same time.

The solo constraint actually became the point: the work is about one person's sustained attention to another. I deliberately avoided grants that required a team because that would change the intimate dynamic. The open-source guide is meant to be used by other solo artists, not scaled.