Skip to main content

Managing active and healthy aging with use of caring service robots

Periodic Reporting for period 2 - MARIO (Managing active and healthy aging with use of caring service robots)

Reporting period: 2016-08-01 to 2018-01-31

Many people with dementia report that dementia affects their confidence to engage in social activities, leaving them lonely and isolated. The MARIO EU project addressed these difficult challenges using service (companion) robots. It involved a team of international researchers, experts, including people with dementia, working collaboratively to develop the companion robot ‘MARIO’.

The key objectives of the project were to address and make progress on the challenging problems of loneliness, isolation and dementia in older persons through multifaceted interventions delivered by service robots; conduct near project length interaction with end users; assist caregivers and physicians in conducting the comprehensive geriatric assessment (CGA); enable MARIO to support and receive “robot applications”; using machine learning techniques and semantic analysis; make MARIO more personable, useful, and accepted by end users; bring the MARIO service robot concepts out of the lab and into industry.

The user design led MARIO robot was deployed across 3 pilots, a hospital (Italy), residential care (Ireland) and community setting (UK). Three phases of testing were undertaken where MARIO engaged with people with dementia in each pilot site. A number of personalised applications were developed including My Music, My Games, My News, My Family and Friends, My Calendar, My Memories and My Chat. In the final phase the impact of MARIO was evaluated using qualitative and quantitative methods. Key stakeholders at each site were involved in this evaluation and included a total of 38 people with dementia, 28 relatives/informal carers, 28 formal carers and 13 managers. The findings revealed that people with dementia are accepting social robots and participants with dementia were happy to have MARIO complete the CGA. Participants with dementia enjoyed engaging with MARIO and it was clear that there was an impact on loneliness as social engagement and connectivity were enhanced. The project has led to advances in semantic data analytics and unique applications tailored to better enhance the connectivity of older people with dementia. However, given the small sample sizes and the limited evaluation period, it is difficult to draw definitive conclusions. Nevertheless, some promising trends as to the positive impact of MARIO on the quality of life, social and cognitive health and loneliness in people with dementia is evident.
Throughout the MARIO project, work has been carried out in line with the Grant Agreement – Annex 1 namely:
-12 WPs completed
-49 Deliverables submitted
-5 Milestones achieved

The consortium demonstrated leadership in both dementia and companion robot spaces in terms of exploitation and dissemination. Key accomplishments include: MARIO Network, commercial partnerships, contributions to EIP-AHA and national/international visibility in terms of news articles, media channels, such as TV, radio and social networks.
The project ambition was built on 6 lines of excellence:
1. Building resilience, rehabilitation and understanding the effects of loneliness, building resilience and rehabilitation for Persons with Dementia
The pilots demonstrate the acceptability of MARIO as a companion robot (e.g. overall system shape and design) and usefulness of its apps to promote social engagement and connectivity and reduce loneliness in PWD (D8.3). At the macro-level, MARIO has been well received as a tool for use by PWD and by both formal and informal caregivers. The project has also advanced the discussion of companion robots as a tool against loneliness for PWD. The socio-economic impacts in communities relate to keeping PWD at home longer as independent community members.In nursing homes, companion robots can improve QOL and reduce loneliness. In hospitals, they can make patients stays less stressful.Business cases are documented in D9.3. Next steps entail progress toward an autonomous companion robot hardware/software platform with unsupervised trials, wider scale pilot activities and use of assessment metrics over longer time periods.

2. Service Robot Enabled Comprehensive Geriatric Assessment and Multidimensional Prognosis Index calculation
Approximately half of the procedure in the CGA were delegated to MARIO and this app has the potential to be commercialised. IRCCS-CNR-R2M are continuing collaboration to exploit this result. The CGA app can save 20 minutes approx. of the 45 minutes typically associated with the CGA. For IRCCS (3-5 CGA per day) this equates to a potential cost savings of €28,000 per year in staff time and only taking the robot about 1-3 hours to do.

3. Frameworks for Ethics and Evidence of Claims
The ethics framework provides a set of innovative practical tools which can be employed in different care robot projects. Outputs will also be published contributing to the state of the art in ethics for social robotics. Future use of these strategies in other studies are likely to facilitate wider societal acceptance of robots in care setting.

4. Cognitive interaction abilities, behavioural and social skills
MARIO incorporates integrated behavioural, cognitive and social skills tailored for PWD that enable it to operate in a dynamical environment to detect, approach, and follow the user and to interact with them through natural language. In combination with MARIO’s applications these functionalities enable the robot to engage with the user, to act in a context aware manner, and to adapt its activity on the basis of the user’ behaviour. The software architecture combines state of the art methods and technologies in an original manner and contributes to progress the state of the art in the area of robotic care and older people.

5. Use of Semantics for language, space and moods recognition
Use of sentiment analysis for adapting the robot’s behaviour to user reaction by detecting emotionally polarised expressions. This ability has been prototyped as a feature of the My Memories app. MARIO also contributes to pushing the state of the art in semantics for social robotics by releasing on the Semantic Web a novel linked data resource, named Framester, covering and interlinking linguistic and factual/ontological knowledge. Framester serves as background knowledge for social robots and includes also data supporting emotion and sentiment analysis. Additional research results that pushes the state of the art include experiments for automatically creating common sense-based background knowledge for social robots. The knowledge needed for supporting daily tasks is missing in the Web of Data hence a set of methods have been developed towards filling this gap.

6. Robot Applications
Developed MARIO apps included My Chat, My Memories, My Music, My Games, My News. One of our unique value proposition aspects is a design process customised for PWD which results in apps customised for PWD. The robot applications are what PWD and caregivers largely interact with – they make the robot useful. This directly impact the uptake of companion robot systems. The consortium recognised the value in progressing toward hardware agnostic and moving forward also making the applications available across various devices. These concepts are discussed in D9.3.