Created by Jasmine Nackash at ITP, The Red Line is a physical device that examines how we adapt to escalating crises, revealing our tendency to shift the threshold of what we consider acceptable over time. It is inspired by the concept of a “red line” — a figurative boundary that, once crossed, signifies a point of no return.
The device continuously gathers news articles from various sources, analyzing and scoring them in real time to reflect the current state of affairs in Israel, Jasmine’s home country. This score is displayed on a live meter, representing the nation’s “livability potential” — ranging from “full potential” to “unlivable”.
A movable red line allows me to set a personal threshold, marking the point at which I would need to take action — whether by leaving the country, engaging in civil disobedience, or reassessing my stance. The line can be adjusted at any time, but each movement is a confrontation: a physical acknowledgment of shifting boundaries and evolving tolerances.
The live score is continuously updated based on real-time news analysis and is derived from a large language model that was trained on Jasmine’s personal values and worldview. It evaluates news articles based on key metrics, including: Contribution to equality, Impact on human rights, Effect on the region’s socio-economic state, and Contribution to peace.
This project serves as both a self-examination and a broader reflection on how individuals and societies recalibrate their moral and political boundaries in response to ongoing crises.
All server-side code was written in Javascript, and is running using Firebase cloud functions, where the project’s continuously updated database is also stored. Real-time news are fetched through Newscatcher API. A fined tuned version of Google AI Studio’s Gemini LLM is used for scoring each news article. The training examples were created manually in a spreadsheet.
Live data is sent over MQTT (using Shiftr.io) to an Arduino that’s connected to WiFi. Arduino libraries used: TMC2209.h, ArduinoMqttClient.h, WiFiNINA.h, Adafruit_NeoPixel.h
Hardware:
Arduino nano 33 IoT
Nema 11 stepper motor
Linear rail guide
TMC2209 motor driver
Neopixel (x1)
Hall Effect sensor (x1)
Custom built circuit
Enclosure uses 5052 Aluminum, cut, partially bent, and powder coated by SendCutSend. Handle was created using 304 stainless steel, cut and bent by SendCutSend with buffing done by hand. Inner parts were designed using Adobe Illustrator / Fusion360, laser cut acrylic and 3D printed PLA.
Project Page | Jasmine Nackash
Project Development Studio | ITP NYU | professor: Pedro Galvao Cesar de Oliveira.
