Second, to this end, we created a responsible innovation ecosystem to further-develop a wearable robot - designed to assist rehabilitation from stroke. The ecosystem established working-relations and common-cause between a stakeholder group of robotics engineers, AHPs and their employer, Blackwood Homes. Engineers found this useful because by bringing together multiple forms of expertise, including experience of living with frailty, it generated immediate insights on their robot design, implemented between workshops. Other benefits included efficient ethics protocol design (clinical experts were embedded on the team), rapid recruitment through Blackwood Homes, and the ability to perform real-time sociotechnical analysis during workshops. Our innovation ecosystem helped us align the needs and capabilities of staff and service users with ongoing design – shifting the evaluation of robots in the real world from binary forms of user-acceptance, to integrated socio-technical alignment. Important because this distributes questions of responsibility to stakeholders across the innovation ecosystem – responsibilities then don’t simply trickle down and add more work to already busy AHPs.  

Third, providing coherence across the project, each of our workshops focussed on aspects of safety training – in our final session we integrated findings into existing curricula for student AHPs, focussing on issues of manual handling. Amongst a pro-technology, British, female participant cohort, we found strong interest in putting prospective training interventions to the test. Future research should look at integrating a similar student cohort directly into the innovation process as it provides an opportunity for rapid design and education prototyping. Future academic papers will elaborate these findings. 

The workshops helped us produce a number of findings.  

Socio-technical findings from hands-on robot testing included: 

• The top attachment causes strain on the upper arm during use. 

• The device needs an emergency-stop button to depressurise 

• In-person testing showed that the device is much lighter than it looks 

• It’s good that it can go over clothes 

• It feels safe to wear once you have it on and have tried it, but it initially looks scary. 

• Opened up conversations about what other uses are there other than stroke rehabilitation? Can it be used for cerebral palsy for example? 

• The importance of video content might enhance other inputs/ways of engaging.  

• The carer should operate the device, not the user. 

• Writing ‘upper’ and ‘lower’ on the two attachments, plus an arrow showing the correct orientation (upward) would help 

• Words were insufficient to convey instructions fully. For instance, the instructions mention an ‘actuator’ but what is this? We found that writing ‘upper’ and ‘lower’ on the two attachments, plus an arrow showing the correct orientation (upward) would help. Also, a red line on the centre of the actuator would help align with the elbow 

• Following this feedback, the instructions were modified to include a QR code link to a video showing the procedure, an image of the device with labelled parts and instructions for removal. A pre-donning checklist was also added, to remind users of items to be checked beforehand. 

• In the teaching and training workshop students really valued the technology, they want better incorporation into training. They want guided supervision through regular updates. 

Eco-system findings included 

• When the group members tried to fit the wearable to one another, it was clear that this was a challenge to complete adequately. This led to a suggestion of the theme of the following workshop, which would revolve around instructions for donning.  

• SR’s point about soft wearable like a blood pressure cuff, instructions on the wearable itself (the advantage here is the instructions can’t be lost and they are right there when someone is unsure about what to do next). Our eco-system approach meant this could be tested immediately 

• Valuable testing multiple technologies beside each other 

• Important to note that the feedback was very varied (interface type, which kind of patient might benefit, the best way to learn about the wearable). The open conversations meant there were many dimensions to follow up. 

Results 

• While many of these findings could have been arrived at through regular user-testing, the ecosystem approach helped accelerate understanding amongst the team. We think it is a promising way forward 

• Our list of human capabilities needed for allied health professionals to work with robots on their terms, derived by members of the team in a previous project, was verified in a new setting – but in diverse configurations. Further work will be needed to show specific details.  

Insights 

• Participants were using heuristics to link talking about robot to something they already knew about, e.g. telepresence robot likened to an iPad, lycra sleeve likened to a support stocking (Robotarium workshop 2). This points to how to deliver training, start with something the people being trained are familiar with using in their role and then move on to the new similar in some ways technology. 

• Responsible innovation eco-systems can include capabilities evaluation. This helps innovation evaluators move from narrow acceptance to situated capability building 

• For Blackwood Homes, the research provided confirmation that learning in the flow of work is preferred method from staff allows us to continue to move away from f2f training removed from work contexts. 

• The best way of having these conversations is not yet codified and standardised. It takes constant reflection amongst the ecosystem members.  

• In terms of education and research: the third workshop at UWE-Bristol made us ask are we setting up the classroom right? We haven’t yet learned how to do this and we’re missing things. This is exacerbated when using robots. One member of the team said It feels like conducing a seance - are you with us, are you there… With a robot this introduces an extra level ambiguity and uncertainty. Our insight is that just because you can be ‘in the room’ doesn’t really mean you can conduct quality observation and research. 

1. At the start of each workshop, participants were presented with short reports of knowledge learned to date in the project. These were actioned immediately in the workshop and represent our deliverable of actionable reports taken forward by participants and ecosystem members.
   
   

2. Further developing the exoskeleton robot, a new PhD has now started working on issues of pronosupination, identified in workshop 2.
   
   

3. Between workshops 1 & 2 we applied for a second ethical approval to expand the cohort of individuals able to try the wearable device. Previously, the inclusion criteria were ‘healthy adults without any neurological or musculoskeletal disorders’ and exclusion criteria were ‘injuries or conditions which restrict the movement of the arm and muscle activation within the arm’. This excluded many of the Blackwood Homes residents so we consulted with Professor Phyo Myint, Clinical Chair in Medicine of Old Age at the School of Medicine, University of Aberdeen.
   
   This allowed an amended and more specific exclusion criteria of: 

• People with peripheral vascular disease. 

• Those with ulcers/wounds on the skin or soft tissue (underlying tissue). 

• Those with active arthritis. 

• Those with peripheral neuropathy. 

• Those with joint deformity.
  
  

4. Aspects of the capabilities framework presented in workshop 2 had been incorporated into Blackwood Homes’ training, teaching and learning competency framework.