Project MM

# WIP

Comment

Hopeject – A Haptic Grief Companion

Inspiration

Challenge: "The Memory Thing"

We were challenged to build a haptic object that connects the digital world with physical space to support people in their personal grief journey. That’s how Hopeject was born – a portmanteau of hope and object.

Hopeject is a customizable digital photo frame enhanced with ambient lighting and an emotional “homecoming” feature. It serves as a comforting anchor – a personal space where memories live on.

We were also inspired to reimagine how people interact with such a device. Instead of using traditional apps, we explored a more human and private approach: managing and uploading memories through a Signal chatbot and an MCP (Memory Control Protocol) Server.

What it does

A story: Isa and her Grandma Gerda

When Isa’s beloved grandmother Gerda passed away, the news left her heartbroken. Gerda had been a world traveler with friends across the globe. Unfortunately, not all could attend the funeral or share their stories in person.

Isa used Hopeject to create a physical memory spot on her shelf. She sent a unique link to Gerda’s friends, inviting them to contribute photos and heartfelt messages.

As these memories arrived, the Hopeject lit up softly, displaying the stories one by one – transforming grief into a gentle glow of remembrance and connection.

How we built it

We combined hardware and software components to bring Hopeject to life:

1. Hardware (Hopeject Device)

  • Raspberry Pi with a TFT display for memory slideshows.
  • ESP32 microcontroller for ambient light and tactile feedback.
  • Raspberry Pi - ESP32 - Connector Derive the color and intensity of the LED by raspberry and controlling the ESP32
  • 3D-printed case reflecting Gerda’s spirit as a globe-trotter

2. Signal Chatbot

  • Enables private, secure memory submission
  • Works seamlessly with both tech-savvy and less experienced users
  • openai-agents for getting results from chatgpt with mcp support
  • signalbot for controlling a signal client

3. Backend

  • PostgreSQL for storing media and metadata
  • REST API for communication between devices, chatbot, and frontend
  • MCP-Server for connecting the signal chatbot with the backend.

4. Frontend

  • Web interface built with Java, Spring Boot, and Vaadin
  • Used primarily for administration and preview (though the chatbot may replace it)

Tech Stack (Containerized Environment)

  • Raspberry Pi (Linux, Python), ESP32 (Arduino)
  • Signal Bot (Python, signalbot, openAI-Agent)
  • MCP-Integration with FastAPI
  • Spring Boot App (Java + Vaadin)
  • PostgreSQL
  • Dockercontainer for Signal Bot, Spring Boot, Postgres, FastAPI

Challenges we ran into

  • As a small team of four, we divided responsibilities across hardware, backend, chatbot, and frontend – requiring tight coordination and flexibility.
  • Networking issues delayed progress on the hardware.
  • Frontend development was a bottleneck, as none of us had deep frontend experience. We iterated through several frameworks before settling on Vaadin.

Accomplishments we’re proud of

  • We built a partially functioning prototype of a grief-supportive IoT device within the hackathon timeline.
  • Our Signal chatbot turned out to be so effective that we considered replacing the traditional web frontend altogether.
  • We created an emotional narrative that connects technology with empathy.

What we learned

  • Building across hardware, backend, and frontend is tough – but incredibly rewarding when it all comes together.
  • Sometimes, minimal user interfaces like chatbots are more intuitive than full-fledged UIs.
  • Combining emotional storytelling with tech can amplify the impact of a project.

What's next for Hopeject

  • ✅ Improve the chatbot to support richer media and voice messages
  • ✅ Add a scheduling feature to display specific memories on certain dates (e.g., birthdays)
  • ✅ Enhance ambient lighting to reflect emotion or mood
  • ✅ Explore NFC or physical tokens for memory activation
  • ✅ Open-source parts of the project for the grief tech community

Built With

  • Raspberry Pi
  • Java
  • Spring Boot
  • Vaadin
  • PostgreSQL
  • Signal
  • ESP32
  • 3D Printing
  • Docker

Team

Daud, Caspar, Jakub, Bastian

Built With

Share this project:

Updates