A $20 clip-on monitor that classifies IV failure modes in real time — built in 36 hours at MedTech Hackathon 2026.

IIIV is a low-cost, multi-sensor attachment for gravity-based IV systems that detects and classifies three failure modes — normal flow, occlusion, and leak/disconnection — through sensor fusion, then streams alerts to a centralized React dashboard for ward-wide monitoring. Designed for understaffed and low-resource clinical environments where existing smart-IV systems are cost-prohibitive.

The name IIIV is a play on "IV," with three I's for the three states the system detects.

📊 Pitch Deck · 🎥 Demo Video · 🧰 Wiring Diagram

• $20 prototype vs. $400–$500+ for commercial competitors (DripAssist, Monidrop) — ~95% cost reduction
• Three-state classification through sensor fusion, addressing the monitoring gap that single-sensor drip monitors can't close
• Clip-on form factor — works with any standard gravity IV bag, no tubing modifications required
• Real-time ward dashboard — single nurse can monitor every IV bag on a floor from one screen
• Offline-capable firmware — runs directly on microcontroller, no internet dependency for the edge device
• End-to-end stack — Arduino/C++ firmware and React frontend, both in this repo

The deck cites two statistics that frame the problem:

• 36% of IV catheter transfusions fail before therapy completion (Marsh et al., 2024)
• 72–99% of clinical alarms are false alarms, driving alarm fatigue (Sendelbach & Funk, 2013)

These numbers describe the same root failure: gravity IV systems have a monitoring gap. Existing devices like Monidrop detect drop rate only, fire threshold-based alarms, and provide no failure classification or sensor redundancy. A nurse hearing an alarm doesn't know whether the bag is empty, occluded, or leaking — so alarms get ignored, and real complications get missed. IIIV closes that gap by classifying the failure mode itself, so alerts are actionable instead of ambient.

Two sensors feed an Arduino-based classifier:

• Water-level sensor (SL067) on the bag → tracks reservoir depletion trend
• Photoresistor on the drip chamber → tracks flow activity via light interruption

Fusing the two signals disambiguates failure modes that are indistinguishable from either sensor alone:

Classified state changes stream to a React dashboard that aggregates patient telemetry across an entire ward. Alerts ping the originating room directly, so medical personnel can respond and triage without parsing which IV setup is failing.

Note: prototype-vs-finished-product comparison; the main point is that the failure-mode classification capability is achievable at a fraction of incumbent BOM.

Firmware notes:

• iiiv_classifier.ino is the production-intent build — analyzes drip rate against fluid level for use with medical-grade IV bags.
• iiiv_classifier_demo.ino is the hackathon prototype — reads water presence as a boolean to accommodate our team's handmade IV bag. Less accurate in a clinical context, but the right call given the sprint constraint.

1. Clone the repo and open the desired .ino file in the Arduino IDE.
2. Wire components per wiring diagram.
3. Select your board and port, then upload.
4. Check out the dashboard demo: cd dashboard && npm install && npm run dev

• Movement ML model — distinguish patient-movement artifacts from true failure signals (gravity IV pumps are sensitive to motion, which is a known false-alarm driver)
• Predictive failure detection — anticipate occlusion onset from upstream drip-rate trends rather than detecting after the fact
• Sensor improvements — higher-quality and longer-range water-level sensor; photoresistor calibration for varied ambient lighting
• Enclosure design — assembly and casing for clinical-environment durability

• William Nguyen
• Ruy Okaji
• Kristina Siju
• Shreshta Kallem
• Krithi Iyer

Won 2nd place in the MedTech Hackathon's Clinical Application Track.

• Couperus, K., Kmiecik, K., & Kang, C. (2019). IV DripAssist: An innovative way to monitor intravenous infusions away from an outlet? Military Medicine, 184(Suppl. 1), 322–325.
• Marsh, N., et al. (2024). Peripheral intravenous catheter infection and failure: A systematic review and meta-analysis. International Journal of Nursing Studies, 151, 104673.
• Sendelbach, S., & Funk, M. (2013). Alarm Fatigue. AACN Advanced Critical Care, 24(4), 378–386.

Full reference list in the pitch deck.