top of page


A concept for a voice-controlled tool to help Nurses prioritise tasks
Problem Statement:

Nurses are understaffed and overworked. Nurses want to have more time with patients. 


I have friends and family who are health professionals. Nurses, occupational therapists, technicians, health care assistants. So I set about to ask nurses and health professionals within my network about the challenges they faced at work.

Getting further information

Having identified a clear area to look into, I reached out to a member of our research team who specialised in the field of healthcare. She was able to answer lots of questions and also give me information about other tools in use as well as additional reference material.

Refining the challenge 

Armed with the answers from my interviews I held a brainstorm with the ID team around how we could replace the existing technology. Collectively we voted that voice assistance could be really promising. I then teamed up with a developer who was keen to help make the POC. 

So now I had


The problem

The parameters

The skills to make POC.

Identifying a problem to pursue 

Yet again it came back to time. Whist talking to one of the interviewees I discovered that one of the challenges this nurse faced on a daily basis on the was prioritisation. The current method for managing prioritisation on a ward is the light board.

Identifying the pain points 

The call bell links to a board and a bed number. The nurse has no idea why the person is pressing the bell they just see the light on for that bed number. This becomes more problematic when there are two lights on as then she must decide whom to help and in what order.


We proposed the following, instead of a bell the patient would use the voice assistant and instead of the light board the nurse would be able to see requests and reasons via her mobile computing device.

Rules for prioritisation

There are some overriding rules in terms of urgency but ultimately nurses know their patients and their patterns. They are making calls based on that knowledge, considering the severity of condition etc. 

The prototype

What we had at the end was a working MVP prototype, using Google Assistant, Dialogflow and Firebase to update the UI. Below is a flow card explaining the journey. 


Below is a screenshot from our project in Dialogflow, we had to input various training phrases and action parameters. 


I no longer have access to the UI as the developer hosted and owned the logins for the project and they left the company. But here are some examples of wireframes for the UI the nurse would see on her device.

What I learned

I learnt how to build a voice prototype and all the work that goes into it, and I really enjoyed the challenge.


What we had at the end was a working MVP prototype to test the concept, it was built by myself and one developer using, Google Assistant, Dialogflow and firebase to update the nurses UI.


I presented the concept and prototype to our global ID team and got good feedback.

The medical sector is challenging due to multiple systems being used within hospitals, as well as all the legal issues around the sharing of data 


We discussed the possibility of the concept being used in many other industries that have a lot less red tape. Imagine if a waiter knew what a table wanted before they got there?


Visibility, in order to prioritise tasks, is a tool that could be adapted to suit the needs of any sector. 


Like what you see?

Let's chat.

bottom of page