Claudia Maw

About

Claudia is a Biomedical Engineer turned healthcare designer who aspires to be a cog in the engine which sees a world in which peoples' health outcomes are supported by empathic, intelligent design and scientific rigour. Her time at the RCA in Innovation Design Engineering has been spent learning how to uncover context and need, and how to provide work that is relevant and beneficial to the world. 

Among other innovations on her RCA journey Claudia has designed a miniaturised defibrillator, worked on a 3D printed prosthesis, chiselled a chair from slate and created a system designed to create value from waste for school children in Nairobi. This variety of works and eye opening series of briefs has given her a broader perspective on design and process. 

In her spare time, Claudia is a costume creator and was named the UK national representative in the annual global contest during her time at the RCA. Though this competition delayed to 2021, this love of making and design has carried her through home-bound times.  

Contact

Degree Details

Statement

"2020 saw me working solely on my major solo project. Apalled by the devastation faced by family friends in the flooding throughout the 2019/2020 winter, I was inspired to look into flood resilience solutions and how they might be empowered to face their recovery. Resolved that I would work in a human centric fashion and put the users first, the output is heavily grounded in interview research performed in East Yorkshire throughout the first few months of the year. 

Though the challenges of remote working took the work away from its human heart, the project is still very much built and designed around the stories and experiences observed in the field. This way of working is something that I have valued immensely and will strive to always bring to every design challenge. I would like to take this chance to thank the inhabitants of Fishlake for their immense accommodation and contribution.

Going forward, I am moving on to work in healthcare design, addressing the need for better diagnostics in a tough time for medicine."

ObjectivEyes

Special thanks to the residents of Fishlake, East Yorkshire.

Flood victims’ lives have already been turned upside down. We give them a tool that’s flexible, honest and quick so they can use it according to their priorities, and focus on what they need.

Research

Background
Background — When a flood hits your home, you lose everything you own. Your walls, your floor, every piece of soft furnishing must be thrown out.
Flood journey
Flood journey — I focused my research in an around East Yorkshire, which has been devastated by flooding in recent years. This mapping represents an aggregate of the learnings obtained from multiple interviews across villagers hit by flooding in the last 20 years in East Yorkshire. From interview tools, I built a journey map of what a typical flood experience looked like, and split it into four phases. Pre-flood where warnings arrive, the flood event itself where water enters the property, the displacement phase wherein a homeowner finds temporary accommodation and the recovery phase. I found this recovery phase most rife with opportunity for design intervention, particularly in the possessions and home insurance space. Some cited the insurance process was worse than the flood itself. Flooded persons feel a real sense of powerlessness and total dependency in their new position, and there is no faith in the insurance system. They may have to wait for months for a loss adjuster to make a first visitation to get the process started; during which they cannot touch anything for fear of losing out in their claim, and objects that were salvageable become mouldy. There are three main tenets I wanted to my solution to meet. Empathy, offering the flexibility a flooded person needs. Empowerment, allowing them to start recovery on day 1. Efficiency - linking all the disparate free floating tools into one solution.
Proposition summary
Proposition summary — Drawing these all together, I resolved to make: ObjectivEyes. This app is designed to enable flood victims to catalogue the water- damaged contents of their home, link them into the recovery services they need, and offer a meaningful contribution in taking their homes back from day 1. This skips the months of waiting for aid, and gives them a start then and there.

How it works

Point your camera into your room...
Point your camera into your room...
... and the app detects your possessions.
... and the app detects your possessions.
Experimental validation
Experimental validation — Computer vision technology has seen use in multiple applications, from sorting recycling to museum cataloguing. I believed this to be a suitable application, and set to learning the necessary skills required to prototype a proof of concept.
... and further experimental validation
... and further experimental validation — I did a series of experiments to gain skills in the area, and to produce a final interactive prototype. While software access in lock-down proved trialling, I worked through online APIs, an Arduino and raspberry pi based camera system, python scripts before eventually settling on Javascript and React apps as the most deployable. I also validated training my own model, and created one that could spot forks. Not eminently useful, but an important step in understanding the process. On the user interface side, I did a series of remote tests with interactive wireframes, and took the feedback from six users and made edits such as altering logic flow, making layouts more visual, less text, and having more concise or obvious home screen options. This went through a few iterations before settling on the concept visualisations.

It’s as simple as, you take an image or video of your room, drawer, or object, and object detection will pick up everything it can. On screen indicators will guide you, and for anything that’s missed, you can input this in manually if the app gets it wrong.

Systems

Product journey map
Product journey map — All in all, the app modifies the overall recovery journey by adding a cyclic and ongoing cycle of communication between homeowner and loss adjuster - without the months of waiting for visitations in between. It saves time, and money, and gives stranded people something to work towards.
System map
System map — This system provides isolated individuals a link to support structures and services they may otherwise be unsure how to reach in one place, and puts them at the center of information flow which runs between firm, homeowner, restoration and building companies and their local communities.
Machine learning system architecture map
Machine learning system architecture map — Object detection protocols rely on the model that is used to train them. This paints an architecture of a system that would be able to operate offline.

Size: -

Try it!

Flood victims’ lives have already been turned upside down.
Flood victims’ lives have already been turned upside down.
We give them a tool that’s flexible, honest and quick so they can use it according to their priorities, and focus on what they need.
We give them a tool that’s flexible, honest and quick so they can use it according to their priorities, and focus on what they need. — Thanks!

I’d like to leave you with my final prototype - you saw it working in the video - a demo web application can be found at. It runs with a Javascript-implemented Tensorflow classifier, which while non ideal, gives you a sense for the versatility of identifying common home items.

https://objectiveyes-webapp.herokuapp.com/

Sponsors