Mobile Social Robot for Healthcare

Title Mobile Social Robot for Healthcare
Summary Pilot study about a small interactive mobile robots for therapy and healthcare in homes.
Author Matthias Mayr
Supervisor Roland Philippsen, Magnus Hållander
Level Bachelor
Status Finished

Project Description

Socially interactive (e.g. Keepon, PaPeRo) and huggable (e.g. Probo) robots have been shown to hold great promise in therapy and healthcare. However, for unsupervised applications in homes, these robots' lack of mobility is a serious drawback. A simple and yet potentially very effective way of addressing this problem is to mount a small socially interactive robot on a small mobile base.

The Halland Health Technology Center (HCH) together with the Halmstad Intelligent Systems Lab (IS-Lab) proposes to conduct a pilot study to demonstrate the usefulness of such a mobile social robot. The outcome will be a demonstrator that can be linked to the intelligent bed developed at HCH. This will show that a mobile robot can be used, for instance, to check on the condition of a person in the bed, to interact with them in order to inquire whether they need help, or to remind them to exercise or take medicine.

Another scenario would take place in an instrumented apartment, such as the one currently developed by CAISR. Here, the mobile social robot could be used as a flexible human-friendly sensing and interaction device for a monitoring system.


Some of the points that need to be considered in the course of this project came out of a discussion between Magnus, Nicholas, and Roland.

  • mobile base: make or buy?
    • reuse our Pie robots? leverage a new platform such as Turtlebot?
    • maybe Roomba educational? probably comes with charging station, would be nice
    • onboard sensors? could include IR detector (easy to find people, may be possible to avoid full-fledged computer vision) </li>
  • control approach
    • teleop? (from intelligent bed computer)
    • onboard?
      • android tablet? would be nice for display, camera, interactivity
      • netbook? would be easy and cheap
      • embedded PC? would be most integrated, can cooperate with CC lab
    • interactive "gizmo"
      • make or buy? question of cost... maybe build something inspired by but simpler than Keepon
      • do a more thorough search for existing platforms first
      • list of features: two eyes, movement, sound, maybe lights, microphone
    • localization?
      • onboard? need e.g. camera, laser scanner, may use beacons (e.g. LEDs on the bed and elsewhere, commanded via domotic system)
      • offboard (invasive vs non-invasive)? could use overhead camera, beacons (spiral patterns: ask Josef, reuse setup of robotics class)
      • maybe not needed at first (single room scenario): can employ line following or other tricks
  • scenarios, constraints and opportunities
    • single room? multi-room? need to cross thresholds...
    • self recharging (standalone demo in the HCH showroom)
    • intelligent bed (monitor life signs of occupant, detect when they get up, ...)
    • instrumented apartment (search for person when environment sensors lost them, ...)