To help people with dementia to live normal lives at home without 24/7 care, hence freeing up hospital beds and allowing people to keep their independence. Hospital bed cost per day: 360pounds. Smartphone: 80 pounds. Sensor: 10 pounds.

What it does

It helps monitor people with dementia. As a consequence, it helps caretakers become more efficient, by:

  • avoiding unnecessary checks on their patients, and
  • knowing which patients require attention, and with which priority

e.g. patients are assigned medicines, and are set intake times. A light sensor in the medicine drawer detects whether the patient has taken the medicine, and it gets reported to the server. It is detected if a patient doesn't take medicine. The patient is called to remind them to take the medicine, and ask why haven't they, if there's a reason. If the patient replies, If the patient seems in distress or doesn't reply the call, an incidence with high priority is generated for the caretakers. The caretakers have a incidences dashboard, and they act upon those incidences.

How we built it

  • MySQL for the database.
  • Python + Flask for the server.
  • JavaScript for controlling Puck.js (bluetooth light sensor).
  • Nexmo to make calls

Challenges we ran into

Having to learn a lot of stuff (request, APIs, programming, MySQL, cookies, Flask, Different Python versions. Code stopping to work because Puck.js crashed, and we didn't know what was wrong. Not being able to install MySQL. Time-constraints.

Debugging More debugging

Accomplishments that we're proud of

Team relationships Getting the light sensor and phone call to work! Learning A LOT.

What we learned

Learnt new coding How to persevere despite setbacks

What's next for Dementia Care

Open platform to add more sensors to the homes, e.g. gps, temperature sensors to check if the oven has been left on, motion sensor to check if they have been using the fridge Personalisation of the voice recorded message, allowing carers to upload their own voice Recording the voice call to check what the patient has responded

Share this project: