EP 117: Designing Open-Source Medical Software | Maya Friedman & Kelly Watson
On today's episode, we are going to talk about designing open source medical software.
Maya Friedman is a digital product design lead and art director who designs solutions for unmet needs in the health tech and femtech spaces. She currently leads the user experience, visual and sonic design for Tidepool's FDA-cleared, automated insulin dosing mobile application, Tidepool Loop. She is also the founder of The Period Project, which supports the creation of an ecosystem of software solutions, research initiatives, and education to address the unmet needs of women+ with diabetes, through the different phases of a woman’s life cycle including menstruation and pregnancy. Prior to joining Tidepool, Maya was a creative lead for Allbodies Health, designing education to improve sex, mind and body literacy. She also worked as a creative at Made Music Studios, driving the design and strategy for digital-audio experiences, most notably leading a partnership with the Cooper Hewitt Design Museum to develop an audio-first design exhibition around alarm fatigue in hospital settings. She thrives leading a team of product designers, illustrators, user researchers and sound designers to design solutions for topics such as health and body literacy, menstrual health and alarm fatigue. Maya has her MFA in Media Design practices with a focus on femtech and sextech design from ArtCenter College of Design.
Kelly Watson is VP, Product & User Experience for Tidepool, a digital health nonprofit building solutions for people with diabetes. Kelly has worked to ensure the patient, the clinician and care partners are at the center of the product design process, leading Tidepool to achieve rapid market adoption growing its global ecosystem to over 450,000 people with diabetes and 12,000 clinicians. At Tidepool, she has led partnerships with many of the world’s largest diabetes medical device companies and technology companies. Previously Kelly worked in bench research at the Center for Biologics Evaluation and Research, FDA (CBER-OVRR DVP) and the National Institute of Mental Health prior to building out a product design consultancy. She is an angel investor and advisor for emerging digital health startups. Kelly advocates for patient ownership of health data, interoperability and open healthcare standards.
Episode mentions and links:
Tidepool loop has received FDA clearance
Maya’s restaurant rec: Saffy’s
Kelly’s restaurant rec: Hook Fish Co.
Follow Maya: LinkedIn | Instagram
Follow Kelly: LinkedIn | Twitter
Episode Website: https://www.designlabpod.com/episodes/117
Episode Reflection
This week’s episode was a personal one for me so I was excited to put on my co-host hat and join the discussion around technology for managing insulin dependent diabetes. I was diagnosed with Type 1 diabetes late in life at the age of 26. I had just graduated pharmacy school and started my first job as a pharmacist at the Children’s Hospital of Philadelphia. After a bout of nasty back-to-back illnesses, something common for people working in a pediatric hospital for the first time, I began to exhibit a group of symptoms that resembled something I had l learned about in my studies. Frequent urination, unquenchable thirst, mood swings, and extreme tiredness. I found myself falling asleep in my chair during the workday and tried to attribute it to needing more sleep. I joke about how I realized something was wrong with me when my parents came to visit and I found myself having to use the bathroom more frequently than my 60-something-year-old mother.
I dug out the glucometer I was given in pharmacy school to learn about diabetes self-care. A right of passage for many pharmacy students learning self-care is to “pretend” they have diabetes for a week, testing their blood sugars and documenting them multiple times a day, to experience how difficult the things we ask our patients to do really are. A quick finger prick confirmed exactly what I feared but deep down already knew was true. My blood glucose was in the 200s. I had diabetes, and thanks to being a pharmacist, I knew exactly what that meant. Later the same month I received my own diagnosis, a close friend from college who had Type 1 diabetes from childhood, Peter Millin, died from complications relating to his diabetes. In my grief, I saw my future laid bare before me, years of struggling with managing this life-altering chronic illness, an early death. Could my story be different from Pete’s?
The answer is YES. Thanks to people and groups like Tidepool, advancements in medicine and technology mean that today, it is more possible than ever to have a long and healthy life with diabetes. When I was diagnosed, my treatment meant finger sticks and multiple daily injections with two kinds of insulin. I was plagued by highs and lows thanks to a lack of information that simply could not replace the functionality my body had lost with my disease. Today, my continuous glucose sensor gives me a 24-hour view of my glucose levels and my insulin pump holds 3 days of insulin and automatically adjusts itself to keep my blood glucose in range. It’s incredible to imagine that in the span of a century, diabetes has gone from a certain death sentence to one where a functional cure is on the horizon. We owe it to those who were born too soon to benefit from the technologies we have today, to keep pushing for a cure. To keep improving the tools that we already have access to and to have the audacity to imagine a future entirely unrecognizable from the reality of today. In this week’s episode, we talked a lot about agency and the power of design to return agency to people. I hope stories like the ones we spoke about in today’s episode will inspire a generation of designers to work in health design and strive to return agency to people using the principle of good human-centered design. We don’t have a cure yet, BUT there is SO MUCH MORE we can do to improve the lives of people suffering from chronic disease TODAY simply by designing better products, services, and spaces and making lives better.
Written by Rob Pugliese
-
Bon Ku: On today's episode, we are going to talk about designing open source medical software. I'm Bon Ku the host of Design Lab, a podcast that explores the intersection of design and health. We have two guests today, Maya Friedman and Kelly Watson. Maya Friedman is a digital product design lead and art director who designed solutions for unmet needs in the health tech and fem tech spaces. Currently she leads the user experience, visual and sonic design for Tidepool's FDA clear automatic insulin dosing, mobile application called Tidepool Loop.
Maya is also the founder of the Period Project. That helps women with diabetes go through their different phases of their life cycle. Prior to joining Tidepool, Maya was a creative lead for All Bodies Health and work as a creative for Manmade Music Studios. She has her MFA and media design practices with a focus both on fem tech and sex tech design from the Art Center College of Design.
Kelly Watson is a VP of Product and User Experience for Tidepool. And she's led their rapid market adoption growing its global ecosystem to over 450,000 people with diabetes and 12 thousand clinicians. She's led partnerships with many of the world's largest diabetes, medical device companies and technology companies.
Previously, Kelly worked in research at the Center for Biologics Evaluation and Research at the FDA and the National Institute of Mental Health. Kelly is also an angel investor and advisor for emerging digital health startups.
Go to our website designlabpod.com. Sign up for our newsletter, each week, our producer, Rob Pugliese will send you his reflection on the podcast, links to related content and you can learn more about the guests. So you don't have to go to the website. You could just get an email right to your inbox.
Please, please, please go to Apple Podcasts and Spotify. Give us five stars on Apple Podcasts you can leave us a review. Follow us, tell someone about the podcasts that reach out to me on social media on Twitter. I can be found at B O N K U on Instagram at D R B O N K U.
Now here's my conversation with Maya and Kelly. Maya and Kelly, welcome to Design Lab. We are excited to have you on the show.
Kelly Watson: Thanks so much Bon
Maya Friedman: very excited to be here.
Bon Ku: this is such a special episode cuz we have two amazing guests and we also have our producer Rob Pugliese joining us cuz he loves the topic
Rob Pugliese: what's up? Yeah, I couldn't miss out on the conversation today. Definitely not.
Bon Ku: So you are Working at an organization called Tidepool, and you created something really cool. Tell us about.
Kelly Watson: Yeah, absolutely. So a little bit of background on Tidepool. So Tidepool is a digital health nonprofit. We were started in 2013, 10 years ago now with this mission to make diabetes data accessible, actionable, and meaningful. And our founder really started this because his daughter was diagnosed with type one diabetes and, he was struggling at the time to get access to data off of his daughter's devices.
He was seen by the best endocrinologist at Stanford and his daughter was using a Medtronic insulin pump at the time and a Dexcom cgm and back in 2013,
Bon Ku: I'm gonna pause right there. Insulin pump cgm, what, what is that? For those who don't know,
Kelly Watson: Yeah. So for people living with type one diabetes, for those living with insulin requiring diabetes as well, continuous glucose monitoring, is essentially a sensor that you wear, continuously for seven to 10 days that, takes readings from your interstitial fluid to determine what your current glucose is.
And insulin pumps deliver insulin underneath the skin as well. You change that every three days. a, it's a device that you wear in order to dose insulin. And at the time when Howard's daughter was diagnosed, this technology was the best technology. There were Medtronic insulin pumps and Dexcom CGMs. Now there are more, than, just those on the market, but. Endocrinologists, were essentially printing these reports, one from Medtronic to see insulin data and another from Dexcom to see the CGM data and holding it up to the light to see CGM data against insulin pump data. And this is like critical.
You need to have access to all of this data to manage your condition. So Tidepool was founded, Howard at the time was vice president of engineering at Amazon. And you know, Felt like this was just not good enough for care for people living with diabetes. So Tidepool was founded based on that initiative. And yeah, we started the Tidepool Loop project. It was born out of this patient led innovation. Glad to talk a little bit more about that as well.
Maya Friedman: Yeah. so building on that, so Tidepool has two different product lines. So as Kelly mentioned, we have a data management platform, which is showing, your retroactive data uploads from the devices that you wear. And specifically today we're gonna be talking a little about the Tidepool Loop platform and the Tidepool Loop platform started out of what is called DIY Loop, which was designed and created by a bunch of very smart people who were unhappy with the availability of technology to manage their diabetes.
So without regulation a community of people started creating an automated insulin dosing system that works on your iPhone.
Bon Ku: Wait, wait, wait. So, so those weren't doctors at all? They were, they were just patients with diabetes.
Maya Friedman: People with diabetes, not doctors. And it, you know, it speaks really interestingly to Bon to the idea that, This is a bit tangential, but that we look at doctors as this sort of the God of the medical industry. And if you actually look at medical textbook language and the adoption of changing language and medical textbooks takes 16 years for new ideas to be updated.
And so if you think about the technology industry, we are able to so much more quickly. Make adjustments to what is needed for patients. And so Tidepool Loop is a perfect example of that. DIY Loop is a perfect example of that, of expertise moving beyond just clinicians, but to patients, to designers, to engineers, to data scientists who can really support, you know, No knock on doctors. I am the daughter of a physician, but I do see there's a lot of room and expertise to shift towards, other me of the industry. So Tidepool Loop really fills and slots in that space. So Tidepool Loop is an app, it lives on your iPhone, and it is an automated insulin dosing system, so it is on your iPhone.
It connects to all of the Bluetooth connected diabetes devices you wear. So as Kelly mentioned, the CGM, your insulin pump, and it pulls all of that information together in a loop and is automatically adjusting the, what is called basal insulin that you are getting in response to what your glucose monitor is, telling you your glucose reading is.
And so, yeah, an incredible piece of technology started by patients themselves.
Rob Pugliese: One of the things that exemplifies this , if you're not familiar with this movement, it just look up #wearenotwaiting. And you see how amazing of a community is out there, really pushing the technology from a patient-centered standpoint. It's really an incredible story.
Kelly Watson: Yeah,
Bon Ku: even, though this is an app, it's a medical device though, right? how, how is that?
Kelly Watson: Yeah. you know, this product is just received clearance in January, so Tidepool took this project forward from this we are not waiting community. I think, you know, one important thing to remember is that 99% of diabetes is self-management, and. The criticality of dosing insulin is really high. So, you know, people rejected this state of technology to create this system themselves.
This community called we are Not Waiting and pushed this project forward. Tidepool decided in 2018 to take this with the intention of working with commercial device manufacturers and with the FDA to push this DIY driven. We are not waiting driven community innovation forward through the regulatory process.
And my partner lives with type one, he was one of the first people in the world using this system back in 2016. And it's, you know, I think the powerful thing about it is that it has changed his life. Having more time in range each day has shifted how he can care for himself and his life. So yeah, it's a med device. Software as a medical device sits on your phone, and is called an interoperable automated glycemic controller, classified by the agency class two medical device.
Bon Ku: So you all actually had to get FDA approval for a smartphone app? Like is this a common thing? Does that, is that done a lot?
Kelly Watson: This is one of the first times for an interoperable, automated glycemic controller. This is the first time a brainer on device. That's compatible on an iPhone and Apple Watch has been brought to clearance. There's certainly some precedents for software as a medical device in general, but this is the first time in the diabetes industry that, this automated insulin dosing system is on an iPhone and Apple Watch.
Bon Ku: What were some design principles that you used when creating this medical device for, you know, people who have chronic additions? This is a chronic disease that people live with for their entire lives.
Maya Friedman: Yeah, when you sign up to be a designer, I don't think that these are some of the constraints that you consider when you're training. So the idea that you're
Bon Ku: You didn't learn this in design school.
Maya Friedman: I, I did not learn in design school. maybe it should be taught in design school, but no one is like, They're like, how can you make it pixel perfect?
No one is like, how can you save someone's life with the right use of color? So there's really a lot that you learn in terms of shifting the principles of good design that you learn in design school and how that manifests when you're designing technology that is, Truly not only giving people life, but saving their lives.
And so I think there's, for Tidepool specifically, because it is a chronic condition, as Kelly mentioned, very high, over 99% of this is self-managed. It is so important that we are not thinking of this as a better safe than story design. It has to be designed with user agency in mind. And so every approach and every decision we make is finding the right balance between is this safe, but are we also giving the user enough information to make a decision on their own how to manage this?
And so it's this constant interplay between safety and agency. so that is certainly one. Certainly one design principle we're thinking about frequently. I think the idea of discretion is also incredibly important. So, I do not live with type one. I am personally connected to my wonderful team that has, family members with type one or have type one themselves.
So I am working and designing with a group of people that are constantly share, sharing personal stories about, their daughter who is a teenager and who is in math class and does not want their phone going off in math class because it is embarrassing. And so how do we design for user discretion? How are we making sure that any alerting that we're doing is, worst case scenario and designed to be quiet or designed to alert with interface rather than audio?
So I think discretion is huge. I think also there's so much data that is, important when you're thinking about designing for someone with type one. There is data coming from your insulin pump. There is data coming from your CGM. There is data coming from your iPhone, and on top of all of that, there is data coming from your entertainment space on your iPhone, your Spotify.
And so how are you also putting the data in context of when it is most necessary. And that is huge guiding principle in how we're thinking about the design of a chronic condition product. Context is hugely important. So those are some, some definite design principles was not taught those in school.
I do think, you know, given the shift towards software as a medical. I could teach a class on this. I think Kelly, we could teach a class on this but yeah, there's, there's a lot of consideration that does not come from the sort of standard tech mentality around efficiency and speed.
It's really about safety. It's about usability. It's about, user agency. It's about discretion. And so your your idea of what the design principal should be when you graduate design school. Very much shift when you start designing for a software company that is designing class two medical devices.
Bon Ku: Why do you decide to use the iPhone or smartphone as a form factor? Why not just create another medical looking device? Why the phone? Because there's so much stuff on the phone, right? Like it's an enter our entertainment device. Email, Instagram, TikTok. You're putting a lifesaving app on your phone. Isn't gonna get crowded out by those other things?
Kelly Watson: Oh, there's so much to say there. I. I think, unlike industry, so that's traditionally how it went. at the time back in 2016 when this system was, was being developed in this, we are not waiting community. The precedent was, you had these big clunky, separate controllers that you had to carry around in your pocket.
And, that was done because it was controllable. Industry, was constrained by regulatory or roadmaps or business needs, trying to control for risk in a device. And you know, unlike industry, this open source version of D I Y Loop was really designed to meet the needs. People living with diabetes and their caregivers that were frustrated by having to carry around these additional devices and having to manage, not having connectivity, having like, Being able to be remotely monitored in that case with our colleague who's managing their daughter remotely.
How do I think about that? So having discretion, being able to deliver something directly on an iPhone. An Apple watch was a design principle to really connect this to the real world, like non-habit to carry around a different device.
Rob Pugliese: It's crazy, the system of insulin pump that I just upgraded from was operating on a Palm Pilot OS system that probably 90% of the people listening don't even know what the heck that is. Right? Like, I've never seen a palm pilot.
It's crazy. The speed at which things change, It's driven by the medical device corporations, not by what patients and families want and are asking for.
Bon Ku: Maya and Kelly, do you have a question for Rob? Because we were talking before we started.
Maya Friedman: Well, you know, I'm in the weeds all day thinking about, you know, this question you asked Bon, which. How do you put a lifesaving app on an iPhone? And that requires knowing a lot about how alerting and the system of an iPhone works because we are having to retrofit an app and all of the needs that we have into iOS 15 soon, iOS 16, iOS 17.
And so while I'm more in the weeds of it when I am speaking with users. It does come up frequently this sort of hate, love relationship with having it on your iPhone because you have this entertainment device that is YouTube, that is Instagram, and then it is also telling you in the same section of alerts that your glucose is low.
And so just context setting and seeing, you know, Rihanna has a new baby. Next to Glucose Low is a really interesting design challenge and we hear a lot of people that have this hate, love relationship. So I would ask Rob, it sounds like you did just switch to Omnipod 5 but how do you feel about having, or how would you feel about having a lifesaving, app that is right next to your Spotify or your Instagram, and could you imagine living like that?
Rob Pugliese: I mean, that's a great question because I remember when I opened my system, I was really excited to scale down to one device because I'm just tired of having so many things in my pockets and carrying around multiple devices. And then I learned that. The system I'd upgraded to was not yet on iPhone.
It was only on Android, which kind of brings up that topic, right? When you don't control the device like the traditional device manufacturers would do. You have to think about versioning, you have to think about what platform, what operating system version you're on.
So yeah, so I'm still carrying around two devices. I don't have it on my phone yet, although I am happy to have a iPhone looking device now, at least an Android device for controlling my, my glucose. And just that upgrade alone was amazing. And I don't know, that's a good question. Now that you bring that up, I don't know how I'm gonna feel about that.
Bon Ku: And Rob, get this on the Apple store right now. Is it app that you could actually download?
Kelly Watson: It's not commercially available yet. We don't have a timeline yet when it'll be in the app store, but that is the intention. It'll be distributed. And available via the app store, so you can download it. You'll have to have a prescription from your provider, in order to use it. So it's a prescription digital therapeutic. So you'll need therapy settings established with your healthcare provider first, but you'll be able to download and go through some educational content first before having that prescription as.
Bon Ku: I mean, this is the future. You could actually download a medical device
to your iPhone. That's crazy.
Rob Pugliese: It's like, yeah, hold on. Let me go on the app store. I gotta download my pancreas.
Maya Friedman: Truly I um, Bon, I remember during the pandemic, I think it was, I think it was you both actually, that ran a series of talks around healthcare in the time of Covid. And I remember there was a physician on the call talking about remote patient monitoring and how difficult it was to work with patients who were not comfortable managing their own conditions, and didn't have the health literacy to use the technologies that were required to check their vitals.
And I remember one of the clinicians saying she would leave packets of like plastic Health kits on her doorstep so patients could come and retrieve them and then she could help support their condition management from afar. And it's just was so wild to me working at Tidepool and realizing Type one diabetics have been doing this for years, and so there's this incredible, infrastructure built by this community about how to think about self-management of a condition, health literacy, learning your, your own technologies, how to train for these types of self-management that I think are really interesting and applicable as we think about, the future of, to your point, technology. And yes, we joke about downloading a pancreas, but this industry shows that people are very equipped, to self manage.
Often, and you know, more than their clinicians about their own condition. And it's incredible to see, wall out of necessity, the ability for us to manage our own bodies and have technology designed to support that. And so I do think Tidepool and type one s that are really incredible precedent for real future of self-management of condition in some way.
Bon Ku: How do you design these alarms? Rob's continuous glucose monitor goes off and it's so loud and I'm thinking, that's so annoying.
Rob Pugliese: Bond. Do you still? I, I share my blood glucose with bond. He's one of my
Maya Friedman: Oh
Bon Ku: it. I, because they kept on alerting and I was like, this is annoying. I, I don't wanna do it. But then,
Rob Pugliese: couldn't even live with it. He couldn't live with it being, being.
Diabetic, even for a little bit.
Bon Ku: and it's loud because it's lifesaving.
Right. So how do you balance those two from being so annoying, this alarm disrupting everything, but you need this alert, this alarm, because it is literally can save someone's life.
Rob Pugliese: Because you've also seen me when I'm really
Bon Ku: Yeah.Yeah. Then it goes off beep. I mean, it's like so loud.
Maya Friedman: Yeah, that, so I would say, Months, years it's taken us to design these systems. there are so many factors that you're considering when you're designing an alert and alarm system for type one. So first off, you're considering three different alerting systems. So you are having to represent alerts from in the type tide pool product alerts from your glucose monitor.
So that is one set of alert you need to manage. You are also receiving information from your hardware, your insulin pump. That is also alerting about different critical issues that may be happening with your pump. And then you also have for Tidepool Loop, an iPhone. So that iPhone has, normal issues like out of range things that you need to alert on, that have to come through the app.
So that's three different systems you're considering. So that is number one. So creating different alert systems for those three systems is very much a dance between the hardware and the software. And so a lot of the alerts that are required come from the hardware, they communicate with the software. And then we have, on top of those three categories, We have our own considerations around designing a system of criticality.
So a system of criticality is not just represented by the audio, it is also how the app itself communicates visually, or via copy about what is happening. So for example, let's imagine in the Tide pool loop product, you receive an alert that your reservoir is empty, so you are out of insulin. That is a critical situation.
So what will happen first is you receive a alert on your iPhone that comes with audio, which I will talk about in a little bit. so the alert comes with audio. You get that on your phone in addition to the audio. You receive a modal. So, as we all know on our iPhones, we receive millions of notifications.
You will receive a critical notification on your iPhone. It will read, via copy what the problem is when you open the app. The system of criticality does not stop there. The way the product is designed is any place that you are looking for your insulin information we'll read. No insulin in giant red letters, so very clearly a visual cue that something is wrong.
You are not receiving insulin. You also have in the product, again, a little hard because we're not looking at it, but a place where you manage both your CGM hardware and your insulin pump hardware. In that management for insulin pump hardware, there is additional levels of criticality. All of the actions you could normally take there are grayed out.
You cannot use them. So we are really directing the user towards the immediate action, which is refill your reservoir with insulin. So that system of criticality, again, is considerate of an audio alert. A modal alert, you know, UI notification, in app, you know, on the home screen. How are we conveying immediate urgency and then really disabling certain features to really encourage someone to take the action they need to take in order to resolve the issue. I'm gonna pause there briefly, but what I also will say is all of this has to happen within the context of the iOS alert system. So most people do not know iOS has four different types of alerts. There are passive alerts. Active alerts, there are time sensitive alerts and there are critical alerts. And so what Tide Pool does is we consider the entire suite of alerts that may come from either your glucose monitor, your insulin pump, or tide pool itself, and we slot each of those alerts into one of the iOS categories so that the iPhone will manage them as defined by the iOS software and.
Very rigorous process. I haven't even touched on how we actually designed the audio for it as well, which I can speak to a little bit. But this system of criticality is so important. And again, this entire system of criticality is so deeply defined by those principles I brought up earlier discretion.
If we do not have to alert with audio, do not alert with audio. If the user has configured a state, for example, which may be dangerous, but the user configured it themselves, we can alert less because we're assuming that they configured it and they know the configuration they made versus a situation like an unexpected hardware malfunction. And so all of this system of criticality is like deeply steeped in these ideas of user agency and alert only when you need to not alert out of better safe than sorry, design.
Bon Ku: so if Rob is scrolling through TikTok and watching some videos and he forgot to eat lunch, and his blood sugar is dropping from like 60 to 50 to 40, like how does that alert system happened? Does he just keep on getting Bings? Does TikTok just shut down on his phone? Like how would that work in that scenario?
Kelly Watson: So this is exactly, you know, type one requires this constant vigilance and we live in this really distracted world, whether or not we have our iPhone in front of us, or whether or not we put our devices to the side. We're in a hypervigilant state always with managing this condition.
And, you know, we are taking over from the pancreas here and trying to navigate how to push on that user agency and give users the right set of alerts and alarms when they need them the most. When you're urgently low though, there's no question we need to inform the user immediately and break through all of the phones native mechanisms to that.
You may turn off your mute switch, you might put your phone away from you. You might have it on vibrate, you might be in do nots disturb mode, because you are, you know, playing on TikTok during a, a meeting, for example. So, we have to manage around all of those different states. And, critical alerts are one way to do that.
On the iPhone we worked with Apple to get, what's called critical alert entitlement, which is a permission that's granted to devices like Tidepool Loop that allow us to break through in situations that are of urgency, like Maya mentioned, and, an urgent low going into the fifties. This is considered a state that we need to inform the user about immediately. No question about it.
Bon Ku: Hmm. is that like the special alert to someone? There's like a weather alert or an amber alert that goes to your phone. Is that, that type of designation?
Maya Friedman: better
Rob Pugliese: Not that
awful
sound.
Bon Ku: Let's talk about the audio. Do you, do
you have like, do you have like cool audios? I know , Maya, you have a background in audio and tell us about how do you design that sound?
Maya Friedman: Yeah, so. Audio is a really interesting category to think about when you are designing for a class two medical device because at the end of the day, the sound should not be disconnected from its purpose. We've seen some incredible advancements in audio and sound design. For example, a a wonderful new evolution is with the audio design in electric vehicles.
We've seen you. we're so used toward the, our device is that the increase in UX design and audio design has been really prolific. But at the end of the day, even with that incredible opportunity of evolved speakers, the sound for a medical device needs to be connected to its purpose. And so there's a lot of things you can consider when improving the audio.
I think, Rob, you may, Be a really strong contender for hating the screaming pod.
Rob Pugliese: Ah, the beeping,
Maya Friedman: there is, there is a lot of, there's a spectrum with which we can make improvements and so some of that spectrum includes thinking about how do you convey criticality with sound? And I think that a lot of people think, Pitch.
A lot of people think incredibly loud. A lot of people think repetition until you want to throw your phone out the window. I think that there's a lot of ways we can think about conveying criticality through audio. In a lot of it in kind of, more thoughtful ways. And so, for example, one of the things that we, we just actually redesigned some alerts and alarms, and one of the things that we really thought about is going back to the system of criticality.
So how can we design a system that has two parts? So for example, if something is impending, let's say a component of your software is going to expire. We would, put out a warning. That warning may be at a level of criticality at about an audio one or two. So it is not going to be incredibly abrasive.
It is not going to adhere to the critical alerts volume that iOS has baked into the device. So it may actually, use the audio volume that a user already has on their phone. And so we've conveyed that warning. The audio is, what I would say is, is palatable. And then that is a way that we have already signaled to the user that there may be something, an event pending soon that will be more critical.
By doing that, we've set the user up, for expecting something that may potentially go wrong. And so the next set of sounds that we convey will be similar and contain the same notes as the warning, but will be longer and slightly, slightly more urgent in the way it's designed. And so it's really thinking about not only just breaking through the noise, but how can we build again, that system of criticality so when a warning happens or a critical event happens, the user may be made aware of it.
Now, that is not in every situation. There are absolutely situations where. The user is not expecting something to go wrong, like an error. And in that way we do design, that doesn't really have as much of a system of criticality that really is thinking about how do we convey classically or sonically what you think of as an error.
And so that in those situations there is, some standard design work that you do around designing audio for error states. But in terms of creating the system of criticality as best we can integrate that into the product, we try and do so, so the user is not shocked, when they receive, you know, receive these alerts.
I think there is absolutely a point where alerting becomes, no longer useful. And at that point, It is doing more harm than good to alert. So it is incredibly important from a design standpoint to not only design the audio thoughtfully, but thinking about when it's going off, to make sure that when the user hears the alert, that they're actually taking the actions that are recommended.
Rob Pugliese: Right. And to give you an example of like what the opposite of the experience you're, you're describing is, with my, my older technology, it would be the same beep no matter what. And it was loud enough probably. probably somebody studied like, all right, somebody with a little bit of mild hearing loss, what's a loud enough beep to make sure they can hear it. If my pump was gonna be empty in four hours, in four hours, but that was in the middle of the night, my pump would just start screaming at me in the middle of the night, even though there was nothing I need to do other than press a button and tell it all right, fine. In four hours. I'll be awake and I will change you.
Bon Ku: Great. And it would continue to beep at me until I hit that button. And there was no opportunity for me to say, I don't want this warning. And I think that's the most for me, having experienced this for years. That's, that's the the most important thing. And I think one of the things that I love about Tidepool and the community you've built is the concept of choice. Right, because, you mentioned balancing between, safety and agency, but safety's great, you have to have safety. That's baseline. But the thing we live with every day is our lack of agency. so Rob, are you, are you going. Tide Pool when it comes out on the app.
Rob Pugliese: Well, I have, you know, I have Tidepool on my phone. I've been a part of the, the community, who shares their data to Tidepool to
Maya Friedman: Oh, thank you.
Kelly Watson:
Rob Pugliese: to enable the research, to progress.
Maya Friedman: Rob, one thing I just wanna comment on is what you just mentioned, like not having built into your products any sort of control. And I know we talked a little bit earlier about, the work with the F D A and one of my favorite stories that I found most interesting was we have, a feature in our device. So if a user has turned off critical alerts, meaning anything that is going to again, break through any settings you have, is going to have a nons selectable volume that we deem or that iOS deem safe enough to hear in a critical situation. We have a warning, so if a user turns those off It says on the device in big red letters, your critical alerts have been disabled.
Tap here to turn them on. What we do not do is we do not block use of the app and the F D A was worried that this creates safety risk, that we're allowing people to continue to use this device even though their critical alert system has been turned off. And what we did is we came back with something again that is an unexpected response. Rather than saying, okay, we're going to block usage of the product while this critical alerts is disabled. We designed a mute alerts functionality, which we thought would be a more powerful feature to give users the ability to turn off their alerts. There are so many situations where you do not want your device screaming, a concert, a meeting with your boss, a podcast recording. And so how can we. How can we design a feature that is going to give that agency so users are not turning off their critical alert? So rather than. Blocking usage. We said, let's design something for better usage and let's design an additional feature. And this was approved by the F D A. And so it's really about this conversation in terms of how can we propose design changes that help users manage the condition better so they're not going to worst case usage of disabling alerts.
Throwing their phone away, not using the product, how it should be used. And so I think that's kind of what you were getting at, but really interesting when you think about the, design of alerts and alarms from that perspective.
Rob Pugliese: That's awesome.
Kelly Watson: there's been, yeah, studies on this from discontinuation of other earlier automated insulin dosing systems where, cause of the frequent alerting or alarming in the systems. over 60% of people ended up just stopping use of something that will give them years back of, you know, helping to support the automation of insulin and making their life easier.
It was actually counterproductive because it was alerting so much that people were discontinuing the use. So designing like that's a huge design challenge to try to create a system. We'll avoid people, discontinuing a product because of that therapy burden. And working with the agency, we were very, we held our ground on what we wanted.
We wanted to support being able for users to let the system know, Hey, listen, this is not a good time. It's the middle of the night. I need to mute alerts right now. I need, you know, for four hours. Some period of time. We need to allow the system to have that user be able to have the agency to make the decision that's appropriate for them and still vibrate, still have different mechanisms to inform the user of the situation.
But, you know, give them some agency in controlling what was critical in the moment. And, you know, working with the agency, they were a great partner with us, like helping us to design the solution that really, held firm to what we believed was right for the user.
Bon Ku: I'm curious to know your personal journey, Kelly and Maya, into this space where you are designing a lifesaving app. How did you get here?
Maya Friedman: it's a great question. Um, yeah, so, you know, I think to start out, I am the child of an artist and a doctor. So if you had a Venn diagram on my birth certificate, I would be plopped right in the middle. So, you know, I think there's that, I think that, I kind of mentioned this a bit earlier, but I have had an unorthodox career path. And I think it all kind of boils down to this idea of when you think about design problems. There is a lot of the problems that we see on the surface. So when you join a tech company for example, there is the ask for efficiency. There is the ask for design, for speed.
There is sort of these obvious, ideas when you go into tech design, what you're going to be doing. And I am, remember this professor that I had who basically said to me, well, what if you design for discretion or what if you designed for joy or what if you designed to slow down? And so encompassing my design career has really been how do I find areas where the design constraints are very out of the ordinary?
So for years I worked doing sound design for products and experiences. So I was very drawn to this idea that how do you tell an incredible story that is both functional and emotional? Just with audio, you do not have visuals. How are you communicating to somebody? And after going to graduate school, Tidepool was an incredible opportunity to say, okay, how do you design for safety?
What about a constraint now that is about risk and safety and some of these ideas that, again, you're not taught in design school, but that really keep you coming back because they're really hard design problems. And so I. That is what has kept me at Tide Pool. I think Tide Pool also, there is something incredibly cool as a designer of working with employees and your team and your staff that are your users.
And so it creates this very close relationship where you get to ask incredibly personal questions of your teammates. And that also creates for a really. Exciting design environment because you have day-to-day access to the lived condition, in a way that doesn't feel distant. And I think for a designer that's almost like anthropology.
Bon Ku: you're in the wild with this condition, and you grow an emotional connection to the people you're working with. And to me that's incredibly motivating to do the work that I do.I love that design journey. How about you Kelly?
Kelly Watson: Yeah, you know, just as Maya said, like for so many of us at Tide Pool Diabetes is personal as well as professional, and you know, our stories are the why behind the work we do. I came to Tide Pool because of a personal connection with diabetes, but started out in bench research, you know, I was 16 and working in university labs.
Started my career off at the F D A, working in H I V vaccine development, and I had this deep desire to make sense of the world around me. My mom lives with serious mental illness, and at the time I had really wanted to deeply understand that and that root, that core of trying to deeply understand problems that were personal to me, took me from, you know, the F D A to N I M H to really look at schizophrenia research at the time, and you know, growing up with these big system challenges.
How do we care for serious mental illness? How like schizophrenia, how do we navigate the shame, the stigma? How do we continue to understand the real needs? I recognize pretty early on that this is a big system design challenge, and that design should be the scientific methods behind really answering them.
And you know, we can talk about hypotheses all the time and about ideas, but design can help give shape to problems so that we can make sense of them and we can start evaluating them in a rational way. And in that time I met my partner Daniel, who lived with type one. And you know, at the time we were both exhausted.
We were just like seven years ago, just exhausted by the day-to-day management of this condition. And so, set out to address the problems in this space. Started looking for what was out there, what was next, you know, trying to get some of that time back for him, for us, and our lives. In the US people that live with type one they will lose 22 healthy years of life, three years lost to treatment and care, and six to complications, and 13 due to shorten life expectancy.
And that's, you know, that's a problem that we needed to solve and I wanted to find a way to support him. So, met this group of people that were working out in this community. This we are now waiting community and met the founders of Tidepool near the same time. So yeah, it's been a wonderful journey, but a kind of a winding one, similar to Maya.
Bon Ku: And we wanna emphasize that Tidepool is a nonprofit organization so you can actually go to their website and donate. This is not a multi-billion dollar company, you know that it needs support. So we'll put the link in our show notes as well. And Kelly, if someone were to come out and visit you, where would you take them out to eat?
Kelly Watson: Oh my gosh, this is such a good question. so I live here in Berkeley, California and in fairly a recent transplant from the East Coast. And I was just sharing this with Maya, like , we both lived in New York for a while before moving out to California and so like trying to find those places that just feel so much like California.
We ended up stumbling on this place called Hook Fish over in Mill Valley. It's on the way to the beach. to get to Stintson Beach, you have to go to this little intersection and there's a little taco stand, this fish taco place and go and grab like drinks and fish tacos and really feel the California vibe.
There's like surfer shop right next door in an equator coffee and this fish taco stand. And it is the best, the best. So that's my recommendation for the Bay Area.
Bon Ku: At restaurant was designed for me. I'm definitely gonna go there. What, what about you, Maya? You are, you are in California, but in Southern California, in LA right?
Maya Friedman: Y. Yeah, so I'm here in the East Side of Los Angeles, also by way of New York, but I will say that my desire for upscale restaurants that give the feeling of an anything could happen night genuinely has traveled with me. So I like, like close rooms, crowded rooms, audio systems that you cannot hear your own self think or speak. and so
Bon Ku: Mm-hmm.
Maya Friedman: guilty. It is. I like, I like a good vibe. I like decor more than the food itself. So I, I have to recommend this place called Saffy's it has opened .Up in East Los Angeles a couple of months ago. It is adjacent to the largest Scientology Center in Los Angeles, which feels like a tourist attraction.
It's literally painted to match the color of the sky. I will say the tint is slightly off, but, they did pretty well. Saffy's is, is Mediterranean. It is advertised as like, eat with your hands, which I do quite like, and has beautiful decor, like very simple. And I have, the designer in me has been very obsessed with this like burnt orange color.
For a while I just bought a bunch of furniture in that color and the restaurant has it as its kitchen accent and so my partner and I went last time and I made him walk into the kitchen with me just to see the shade of red and he did not like it. so yeah, really drawn to, to beautiful interiors. So Saffy, if you come to La Bon,
Bon Ku: I love. well this is so great. I, love the work that you all are doing. Thank you for working on, on this, lifesaving medical device
Rob Pugliese: Yeah, thanks for continuing to keep diabetes care moving forward.
Kelly Watson: We are privileged. It's been an honor.
Maya Friedman: Yeah. How fun. Thank you so much.
Bon Ku: You can find Kelly on Twitter at K E L L Y I S W O R K I N G. And you can find Maya on Instagram at M A Y A C F R I E D M A N. Rob Pugliese produces Design Lab, editing is done by Fernando Quieroz, Emmanuel Houston created our theme music and the cover design is by Eden Lew. See you next week.
