Changes for page Robotic Partner

Last modified by Laura Ottevanger on 2022/04/05 13:45

From version Icon 18.1 Icon
edited by Tim Huisman
on 2022/04/04 18:18
Change comment: There is no comment for this version
To version Icon 19.1 Icon
edited by Tim Huisman
on 2022/04/04 18:20
Change comment: There is no comment for this version

Summary

Details

Icon Page properties
Content
... ... @@ -3,14 +3,16 @@
3 3  Next to the practical reasons for which we chose [[MiRo>>https://www.miro-e.com/robot]] specified in [[Humanoid Robot>>https://xwiki.ewi.tudelft.nl/xwiki/wiki/sce2022group02/view/Foundation/Technology/Humanoid%20Robot]], we also chose this robot because its dog-like properties support Patients with Dementia (PwD) to recall a collection of memories about a time where things went slower and were less complicated. It emphasizes on the PwD's need to connect with nature and themselves in peace [[(Han et al., 2015, p. 120)>>https://journals.sagepub.com/doi/10.1177/1533317515598857]]. Although the MiRo has a limited range of expression, its limitations also make it more accessible for PwD. PwD tend to perform better with simple social interactions and with items that are similar to what they were used to before the onset.
4 4  
5 5  ==== Other dog-like robots ====
6 -A feature that we like to get from other robots that we would use for a new prototype would be legs that would be able to go up steps. An example of such a robot is the Boston Dynamics Robot [[Spot>>https://www.bostondynamics.com/products/spot]]. It's legs and sophisticated stability functions allow it to traverse difficult terrain, such as slopes and height differences.
6 +Although MiRo was the most desirable choice of the available robots, it is not perfect for our use case. MiRo is frail robot that cannot withstand any handling rougher than soft touches. As described in functions, we ideally would want a physical connection to the dog so that the PwD does not lose focus when walking with MiRo. However, given the frailty of MiRo, a physical attachment would likely result in damaging a €2.400 robot. Next to this, MiRo can only be used inside as it cannot go up steps or walk on any rough terrain. With the right robot, PwD could also take walks outside for example. The final feature that we want to implement is Snoezelen with the robot. Our ideas for this are outlined in the Future work section of the [[conclusion>>https://xwiki.ewi.tudelft.nl/xwiki/wiki/sce2022group02/view/Main/Conclusion/]]
7 7  
8 -AIBO: AIBO is a doglike robot that acts like a dog. [[(Fujita M., 2001)>>https://journals-sagepub-com/doi/abs/10.1177/02783640122068092]]. This dog acts like a real dog, one needs to teach him commands and he will grow up like a regular dog. When the dog is petted, it will respond like a real dog. Features from this dog that are relevant are that it is able to listen to a name and that it will respond like a real dog. Unfortunately, this dog is not able to walk for a long time and is very slow when walking.
8 +An example of a robot that could handle rough terrain, is the Boston Dynamics Robot [[Spot>>https://www.bostondynamics.com/products/spot]]. Its legs and sophisticated stability functions allow it to traverse difficult terrain, such as slopes and height differences.
9 9  
10 +Another, very different example of a dog-like robot, is the AIBO. AIBO is a doglike robot that acts like a dog. [[(Fujita M., 2001)>>https://journals-sagepub-com/doi/abs/10.1177/02783640122068092]]. This dog acts like a real dog, one needs to teach him commands and he will grow up like a regular dog. When the dog is petted, it will respond like a real dog. Features from this dog that are relevant are that it is able to listen to a name and that it will respond like a real dog. Unfortunately, this dog is not able to walk for a long time and is very slow when walking.
11 +
10 10  Another robot that might be of interest is the Guide Dog robot [[(Tachi et al., 1985)>>http://files.tachilab.org/publications/paper1900/tachi1985MIT.pdf]]. This is quite an old project where a robot was used as a guide dog. The problems encountered were mainly with object detection and how the information should be displayed to a blind user. As shown in the survey by Zou [[(Zou et al., 2019)>>https://arxiv.org/pdf/1905.05055.pdf]] object detection improved quite a lot. In the case with people with dementia, we do not mainly focus on blind people and are not trying to be a guide dog, which mitigates the information displaying problem.
11 11  
12 12  The commercially available Joy for All companion pets include cats and dogs. They are far more affordable than more advanced robots on the market and are therefore more easily introduced in a real care home. [[Thunberg et al., 2022>>https://doi-org.tudelft.idm.oclc.org/10.1007/978-3-030-62056-1_51]]. These commercially available robots are usually focussed on Snoezelen and letting PwD pet them. They have a very limited amount of movement. Some of these robots include: Paro Robot which is a seal-like robot which can move its head and tail, the NeCoRo and justtocat which are cat-like fluffy robots which can lay on ones lap and finally the cuddler which is a teddy bear robot.
13 13  
14 -Although MiRo was the most desirable choice of the available robots, it is not perfect for our use case. MiRo is frail robot that cannot withstand any handling rougher than soft touches. As described in functions, we ideally would want a physical connection to the dog so that the PwD does not lose focus when walking with MiRo. However, given the frailty of MiRo, a physical attachment would likely result in damaging a €2.400 robot. Next to this, MiRo can only be used inside as it cannot go up steps or walk on any rough terrain. With the right robot, PwD could also take walks outside for example. The final feature that we want to implement is Snoezelen with the robot. Our ideas for this are outlined in the Future work section of the [[conclusion>>https://xwiki.ewi.tudelft.nl/xwiki/wiki/sce2022group02/view/Main/Conclusion/]]
15 15  
16 16  
18 +