Social Care Robot Challenge 2017 is a national cooperative venture
To begin addressing the wider challenge we will bring together UK experts in social care robotics from both academia and industry to advance our knowledge of how robots can be integrated into the healthcare services of the future. Thus we will start to address the predicted steeply rising costs and strain of healthcare provision and services in the UK. We will focus on a selected subset of the overall challenge of how robotics can be integrated into multiple components of our working healthcare system in order to work with and support health professionals in providing improved physical and cognitive deficit support. To achieve this, we will utilise the expertise and advanced equipment provision in ‘living laboratory’ facilities that exist across the UK Robotics and Autonomous Systems community, where prototype robots can be evaluated in home-like environments. Although there will be many issues to be addressed within and beyond the scope of the 2017 Challenge, it could be summed up as being driven by a motivation to create an architecture for social cognition in care robotics. A vital aspect of this activity will be to allow robots in different Institutions to pool knowledge that they gain from their local interactions and thereby become “more than the sum of their parts”.
Organisation, Dates and Location of the Challenge
This is a very large field of research. Consequently, we must focus on a narrower focus for this activity that will run for the first six months of 2017. A topic, or interlinked set of topics, will be chosen that maximises benefit for all and is achievable in the timeframe, equipment and distributed infrastructure available within the participating partnership. This process will be driven by one or more themed workshops to be held during Spring 2017. Collaborative work will then proceed during for the remaining time until June 2017 on the selected activity focus within social cognition. We will work towards a public oriented event that demonstrates our collaborative findings, which will be held during the UK Robotics Week, 24-30th June 2017. Social Care Challenge 2017 will be a distributed activity held across the ‘living laboratory’ facilities that exist across the UK Robotics and Autonomous Systems community. Participating partners without such facilities will be encouraged to take part by arranging to access those of partners possessing them. Organisations that are not already part of the UK RAS community, but have suitable facilities or track record, will be encouraged to take part.
Research issues to be addressed from a robot social cognition perspective could include one or more of the following:
1. Assisted mobility
2. Personal hygiene
3. Social support
4. Preventative and rehabilitation monitoring
5. Remote assistance
6. Food preparation
Some underpinning questions for the chosen sub-theme to choose from are:
1. Can a robot faced with a new and complex task, get help and gain experience from other remote agents, robot or human?
2. What type of cyber-secure networked communication and control architectures are needed?
3. How are robots best integrated into the digital support package?
4. How can we ensure the robustness and safety of systems during normal operation and that they degrade gracefully?
5. Can we build robots that can work effectively in human-robot teams so that the team overall becomes better than the sum of its individual capabilities?
Underpinning robot social cognition technologies to be advanced could include one or more of the following:
1. Long-term autonomy
2. Social and physical intelligence
3. Spoken language understanding
4. Multi-modal perceptual learning for cooperative control via Learning-from-Demonstration
5. Acquired knowledge transfer between robots, including cyber-security of that transfer.
6. Networked hardware for Human-in-the-Loop simulation
7. Verification and validation of safety, including Assisting in regulation formation
8. Dynamic autonomy
9. Immersive tele-presence
Exemplar scenarios
Facilitated by one or more themed workshops, we will draw on the issues, questions and technologies listed above to identify and design a chosen pilot study, or a small set of interlinked pilot studies, in realistic healthcare settings that promote integration between partners and permit greater user and stakeholder involvement in robot design and development. Assistive robots that have been designed to support independent living will need to work safely with and alongside people, requiring advances in multiple scientific and engineering domains, including artificial intelligence, mechatronics, materials, sensing, and human-robot interaction
An example of how the Social Care Challenge activity might proceed:
(i) Identify a small number of robot platforms that many RAS network partners have or would like to have. Platforms that we might use: Nao, Pepper, Fetch, TIAGo, iCub, MiRo.
(ii) At an initial workshop, partners show what their capabilities are on the chosen platform, or similar platforms, and what further competence they would like to create in the next 3-6 months. A capability might be verbal and non-verbal communication, scene understanding, person recognition, object handling (e.g. in the context of a ‘fetch-and-carry’ scenario).
(iii) Agree on a social care scenario that could be achieved on the robot(s), based on the combined set of competences.
(iv) From Easter 2017, partners start to share code and experience.
(v) Organise an integration week one month before UK-Robotics week.
(vi) Researchers come together to combine competences and integrate into an overall scenario.
(vii) Further refinement and tuning over subsequent weeks including testing scenarios in place.
(viii) The scenario may be demonstrated at a single site or at multiple partner institutions during UK robotics week on 24-30th June 2017.