Embedded - 135: No More Yoda Heads

Episode Date: January 21, 2016

Anh Bui, Vice President of @Benetech Labs, joined us to discuss using technology for good.  Benetech is most widely known for Bookshare, an online library for people with print disabilities. Note that this is only open to people with print disabilities per the Chafee Amendment (copyright exceptions with cause). There are some public domain books you can search through on the site. Martus is another of Benetech's core programs, in their human rights and civil liberties program. It is an open source, secure information collection and management tool. Poet Image Description Tool is a Benetech tool to aid in making visuals more accessible to everyone. Some accessibility guidelines and techniques: W3 Web design accessibility Apple's many accessibility resources POUR (Perceivable, Operable, Understandable, Robust) website design article Enabling The Future is the group that 3D prints prosthetic limbs.The Dean Kamen water filtration system is called Slingshot. Hackaday Prize (2016 announcement is coming!) For more information about the embedded software position at Avid Identification Systems, please email Mark (Engineering Manager) and CC Karen (HR Manager). 

Welcome to Embedded FM. I'm Alicia White, here with Christopher White. Our guest this week is Anne Bowie, Vice President of Benetech Labs. But before we get started with her, I've got a little job announcement from a manager who is an Embedded FM listener. Avid Identification Systems is a Southern California company that has been saving pets and reunited families for 27 years with their products and services. They're designing a new generation of RF ID tags. They're looking for an embedded software engineer to help create a new product. Their ideal candidate would have experience with digital signal processing, software-defined radio, as well as experience with USB or Bluetooth.

Seems like a pretty high bar if you've got two of those, maybe drop them a line. I will put their contact information in the show notes. And Avid is located in Southern California, the Inland Empire, which is where I grew up, in the city of Norco, which is not the city I grew up in. Okay, now let's get started. Hi Anne, thanks for joining us today. Thank you. It's great to be here. Could you tell us about yourself? Sure. I think, as you said in the intro, I'm the VP of Benetech Labs, where I get the fun job of overseeing the team that experiments and prototypes new technology for pressing social needs.

And we are the experimentation and prototyping innovation arm of Benetech, which is a tech nonprofit that focuses on working with communities in need by creating scalable tech solutions for them. And the work that we do includes technology for human rights and for accessibility in education and information access. I expect we have a lot to talk about. But we've been doing this thing where we ask you lightning round questions, and the goal is one-word answers. And I guess the first one will be consonants or vowels. Vowels.

Introvert or extrovert? Introvert. Favorite thing to listen to on your commute npr mac or pc both wow wow have you seen the new star wars movie yes i don't know if i should ask a follow-up to that go ahead uh bb8 or r2d2 r2 actually i can't choose that's not a one-word answer but both that's a one-word answer well yeah okay i'd accept that one uh gryffindor or Ravenclaw? Okay. Manager, mentor, or coach? Mentor.

Mercury or Mars? Mercury. And one last one. Newspapers, magazines, Twitter, or an RSS feed? Ooh. None? None? None or all, I think is a better... that's a tough one.

Where do you get your news from? Everything. Fair enough. Okay, So Benetech, if no one has, if, if I haven't heard of them before, what, what would I most interact with Benetech? I mean, how would I get introduced to them if I'm just an engineer going about their daily business? Well, so we have two flagship products that are probably the, we're best known for. So the first is Bookshare, which is the world's largest online library

of accessible digital books for people with print disabilities, such as blindness, dyslexia, any kind of disability that will prevent you from holding and using a print book. So Bookshare goes out to more than 370,000 people around the world, largely students in the United States, and it contains more than 350,000 books. So it's our broadest reaching product. And many of the people who come to us come to us because they've heard of Bookshare. And it's robotic voices, mostly?

Actually, so the way Bookshare works is we have a collection of texts in a variety of different formats there's a there's a basic uh text uh basically an xml based kind of format um that we house we store the books in and then we convert them into a variety of other formats including braille and also audio formats that have a uh text-to-speech voice that provides the audio. But it's not something that I can sign up for. You have to have a doctor's note? Well, you can't sign up for Bookshare to get access to the full collection, correct. You have to have a proof of disability, which is, yes, something you can get from a doctor or any number of other authorized entities that can give you that proof of disability. But we do have public domain content on Bookshare.

So anybody looking to just explore Bookshare and kind of get a feel for, you know, our platform can come in and, you know, read twain or jane austen or whatever have you what other uh public domain content we have there okay so for the for the other content why why can't i get to all of it uh so we are able to make all of those books available uh in the united states to people with print disabilities thanks to an exception in the United States to people with print disabilities, thanks to an exception in the copyright law called the Chafee Amendment. And Chafee is very specific about authorized entities being able to offer books to people with specific disabilities. And in order to align ourselves with the legal qualifications there, we need to make sure that the people who are accessing all of the content do have the qualifying disabilities. Fortunately, you know, over time, publishers have been more open to giving us digital content so we don't have to go out and, you know, buy a print book, chop it, scan it. We still have to do that for a fair number of books. But for many, many books, the vast majority of our books, don't have a copyright exemption, we're counting on

those publishers to provide us licensing for that content, again, for specific audiences. So, that's why there are those limitations. Is that amendment limited to books or other media as well? So, you know, other media are accepted. So, for instance, we can't put plays in Bookshare because drama is not. That's kind of what I was thinking. Okay. Yeah. It does not qualify under Chafee. Huh.

Yeah, it's a complex. I mean, I think the community is doing a lot of work right now trying to update the law and trying to actually, there, there is something called the Marrakesh Treaty, um, uh, that we're trying to get ratified in, uh, basically globally, um, that would allow these books, uh, for people with print disabilities, allow these books to be, uh, uh, distributed globally or distributed in all the countries that have ratified the Marrakesh Treaty. So, you know, I think there's, there's good news on the horizon for people with disabilities being able to get access to more and more content. But in the meantime, there are some limitations on what we can distribute. So how did Bookshare come about?

I mean, I can see how this is important and how it can be useful to a lot of people but how did it start so our founder jim fructiman uh who uh is well known in the social entrepreneurship field he's a macarthur fellow he he and you know started his career as a rocket scientist, literally. He built reading machines for the blind, which is to say he built text-to-speech machines that worked using core OCR technology. And when he decided that he wanted to, well, let's put it this way. He heard about something you might know of, which was Napster, I think from his son, actually.

And he thought, you know, what a great idea. What if instead of just sharing music, people who are blind and who spend a lot of effort in basically making books accessible so that they can read them could share that work so that they weren't having to do duplicative work and more content could be available to more people. And from that concept, Bookshare was born. What was the other, uh, flagship program? So our other, uh, program is in human rights and our flagship product there is something called Martis. So let me back up a little bit. Our human rights work is focused largely on the collection, storage, and management highly secure system for collecting that data, storing that data in a way where even we don't have access to it. So even if we were subpoenaed by the government to turn over information that was stored in Martis, we would not have access to it, and then do some light assessment of that data.

And we work with groups around the world, usually grassroots human rights organizations, to basically collect stories of atrocities and help them in their advocacy work. There are a lot of times when there are atrocities local to a place and without corroborating evidence, it's hard to show trends. Is that what it's for or something else? Yeah, it's partly, so actually our partners use it for a variety of different reasons. But yes, a lot of it is not just corroborating evidence, but just actually collecting information to begin with about things that are happening. have been working for many years with pen and paper, who have, are, you know, risking their own personal safety to talk to people who are vulnerable, you know, to vulnerable people, or people who have been victims of atrocities. It's a way for them to basically easily enter a digital space,

which is safe and allows them to share that data later on with other groups and with other agencies. How can you possibly make that safe? How can you make it secure enough? Well, that's a great question. And, you know, I think we always, when we say the word secure, we always put a little asterisk against it because there, you know, there is no such thing as absolute security. Not when you get humans involved. Exactly. And at the end of the day, you know, it a personal physical safety is that's not something that a um a secure server is going to help you with um you know there's you know one of the things that we hear a lot about is the scenario where someone is you know it's all it's one thing to have your entire device

encrypted or password protected or what have you it's another thing when have your entire device encrypted or password protected or what have you. It's another thing when you are faced with a personal threat or personal risk if you don't decrypt it for somebody. So you're absolutely right. There is no such thing as 100% safe and secure. But we try to make it that much easier for groups, particularly groups that are not technically savvy and don't have a lot of technical capacity, for those groups at least to have the level of security that many others with that technical capacity enjoy. The other thing I've heard of before we talked was 3d printers in libraries and i understand that you're involved uh with that yes and this takes us full circle back to our star

wars reference um because um one of the the sort of catchphrases for our 3D printer work is no more Yoda heads, which is no offense to Yoda. We all love Yoda. But the idea behind this really came out of our work in our global literacy program, which is the program that focuses on accessibility and increasingly on, or rather I should say overwhelmingly on the education and accessibility. One of the things that we have found in running Bookshare, of course, is that in many cases the problem of text accessibility is pretty much solved, right? There's text-to-speech engines, there's scanning, there's OCR. Text is a pretty clear, has a pretty clear pathway to accessibility. What is also true, though, is that image content, visual content,

is still a largely unsolved problem in that it is highly, highly resource intensive to make it accessible at this point. And even if you had the resources, the different ways that people approach visual content, the different modalities that make sense for conveying the information in visual content can be wide ranging. So for instance, I'm gonna give you a concrete example because content can be wide-ranging.

So for instance, I'm going to give you a concrete example because that'll be easier to grok. For instance, if I'm looking at a saddle curve and I'm trying to describe a saddle curve to someone who's blind, that's a really difficult thing to describe verbally. So the next thing people often think about is, well, okay, if I can't describe it verbally, I can produce a tactile version of it. So, and in the accessibility world, tactile version is very often a raised 2D or two and a half D version of something. Well, as you can imagine, touching a raised 2D version of a saddle curve still doesn't really tell you what a saddle curve is. It's still pretty hard to imagine. So now you're left with, okay that is 3D. So of course, with the advent of 3D printing, consumer 3D printing, and increasingly 3D printing that is,

I mean, I wouldn't say that 3D printing is a completely consumer-ready process, but it's increasingly available to convey complex spatial information in a low volume and quick turnaround way. as we were looking around the community for 3D printing resources, it became clear that many public libraries, particularly in areas where there are strong technology ties like the Silicon Valley, were investing in makerspaces. And makerspaces, of course, are a lot more than just 3D printers on a desk. You know, making is a rich and full, wide-ranging body of practice. But they were investing also in 3D printers and getting funding from the Institute of Museum and Library Services, the other foundations and institutions interested in making, to put, they were getting funding to put 3D printers in these spaces, and they were getting a fair amount of interest from the community. But they were running out of things to meaningfully print.

Yoda heads. Exactly. And we come full circle. And that quote actually comes directly from a museum person that we worked with. He just said, you know, I'm kind of tired of printing Yoda heads. Now, there have been great projects out there that have really taken 3D printing and put a social spin on what can be created. I'm thinking of the Enable group where they're creating prosthetics for children, prosthetic hands, prosthetic limbs for children using 3D printers. They're terrific examples

of 3D printing being used for good. And we just felt this was another place where we could connect teachers and parents with libraries, all of whom have an educational mission to allow them to use these resources to create accessible materials for the classroom. I really like the idea of science and math things that you can touch because images are not enough. I think on your website, I saw the library example being used with an ear canal and looking at the bones inside of an ear on a piece of paper on 2d is so much different than printing out an ear canal cutting it in half and looking at it well yeah even outside of the accessibility uh issue absolutely and i mean you guys just hit the nail on the head what we know, our goal is not just to create a more accessible

classroom for students with disabilities. It's to create a more accessible classroom for all students who may have a better experience if they have, they are able to tackle these complex objects and complex concepts in, you know, a different modality. So, you know, I may be very verbal, and I might do great if somebody reads a paragraph about the ear describing it to me, or, you know, the ear canal describing to me, or I may be someone who really needs to touch the inner ear to really understand how the components are put together. So, and, you know, in neither of those situations may i be diagnosed with a disability of any kind so um you know it's interesting we uh when we started doing 3d work we got a lot of attention from from you know great wonderful attention from groups who um uh are you know

helping kids make um and so we had a group come in from that was a first lego league i think um team and it was a group of six 12 year olds and they came in and we gave them a little talk about you know the work that we're doing and and handed out that inner ear model and you just it was great just to see them light up and really just you know and one of them just turned to the other and said, man, we could really use this. And, you know, I think that kind of aha moment, which we've had not just with them, but, you know, of course, also with people with disabilities, is just so powerful that it just makes all the sense in the world to try to make those models available.

And maybe 3D printing is not the end-all, be-all technology for that. We're a lab. We're completely open to a different technological solution. But it's clear that being able to touch and feel a 3D model is just incredibly important. Part of me is like, duh, of course it is. And the other part of me is like, I totally need a 3D printer now just so I can look at ear canals. You don't because your library may have one.

That's true. And I do live Silicon Valley-ish, so I can probably find somebody. But Benetech Labs, so we've got three pretty disparate things here. We've got an initiative to put 3D printers in libraries. We've got text-to-speech and Bookshare, which definitely got a database behind it. And we've got the Martis with its security aspect. It's got to be a research project on its own. What does Benetech Labs do? I mean, you just research everything? That is a great, great question. So the two products that I described, Bookshare and Martyrs, aren't technically Benetech Labs projects. And the reason for that is because they've graduated. Basically, they're now fully operational products that were taken to

scale. Benetech Labs is about looking for that next product to take to scale. And to do that, we are looking, you're absolutely right, we could just boil the ocean, we could, you know, just, you know, throw our arms out wide and say, you know, at all comers, let's do this. But, you know, of course, what we prefer to do and what we're aiming to do is really look at projects that leverage the expertise and capacity that we have at Benetech already. And that may, you know, there's maybe one or two degrees removed from existing work. So that means, for instance, that, you know, we don't have expertise in embedded systems, our next project is probably not going to be something that's going to require that expertise. We do have some expertise in data management and security. And so, for instance, one of our projects builds on top of that expertise,

which is a secure, right now, it's inelegantly called Secure App Generator. You know, I'm happy to take recommendations for a new name. And the idea of which is, you know, we were getting requests not just from human rights organizations, but other NGOs for a way to quickly set up a system for collecting the data that they need for day-to-day use and storing it securely. And so the Sec app generator is, you know, sort of the next generation of our work in this space. And it allows a non-technical NGO to, within minutes, create an app that collects data and then stores it securely in our servers. And we make that, you know, we're going to make that freely available to the NGOs that need it. And that's work that, again, is not specifically in human rights. And we're talking to groups that work in the environment, groups that work in financial inclusion.

But it's related in terms of its features and functionality. Do you worry about being a target of cyber attack and things like that from hostile states and things that don't agree with what you're doing? Yes, in the sense that that's, you know, it's exactly that concern that our users have, which is why we have the system that we have. As far as we know, we have not been attacked. That's not an invitation. That is not an invitation. I don't think they listen to this show. And, you know, we also, this is, you know, perhaps to our detriment, but I think we have been very careful not to collect information about who uses the service. Right. all the organizations that use artists because we, you know, want to stay, we want to be as,

that's what I'm looking for. We want to protect identities as much as we can and, and pose as little risk as possible for those users. Right, because if that data isn't there, there's nothing to steal in some sense. Exactly. So some of these things sound sort of like last year's Hackaday Prize, which I gather you did talk to those folks a little bit. I did. I was super interested in the satellite work by the way

yeah very interesting well and last year's winner was the iDrivematic which technologically wasn't the most impressive project ever but two guys uh bound to wheelchairs and and very little movement available to them, created this system that allowed them to use their eye tracking software to also drive their wheelchairs and opened that whole project up so that anyone could do it with a very minimum amount of 3D printing and a little bit of modifications. Since they didn't own their wheelchairs, it had to all be add-on, which was an additional interesting challenge.

Oh, wow. I had not seen that. That is super interesting. Well, I do actually have heard what the theme is this year, and I'm not going to give it away, although I can say that it's still do-gooder focused. Oh, not super weapons. Not super weapons, but thank you. You know a lot about the different technologies that are out there.

What devices could people build at home or in small groups, small scale stuff that can make it different, that can help with, what did you say, solving pressing needs? So let me take a step back from that question a little bit and approach it from a slightly different angle. It is sort of an impossible question, so it's fine. Well, I mean, because as I think through that question, I think that the biggest hurdle is it's not that people have a shortage of good ideas, right? And certainly not a shortage of great technology and great technological ideas, I think that the danger is always there's a risk in building something that no one will ever use. Oh, yes, yes.

At the end of the day, you had this really cool idea to, you know, mount a camera on X in order to automatically and use image recognition machine learning to automatically read the street signs for navigation. And isn't this great? And it very well could be great. But I think, you know, one of the things that we have learned, you know, in the hard way in some cases is, and is, is that it's so important to need find and really embed yourself in a community, um, that, that has the expressed need before you build anything. Right. Um, and, and so, you know, when, when we look at the operating principles of labs, and I should say labs is pretty new. I mean, as a formal, you know, Benetech is always innovated, and we've always, you know, looked for new ideas to take to scale. But as a formal entity, labs is, you know, just a year old, actually, as of next week. So it's pretty new. And, you know, as we were thinking through our

operating principles last year, it became very clear to us that one of our key, key principles was building community and really doing what we call, exercising what we call ecosystem engagement, where we really go in there and not just talk to the end users, but really talk to everybody around that space, including other technologists in that space. So, you know, rather than urge people to build on their own, which I wouldn't, you know, I'm not trying to discourage people from doing that. I also think tinkering is awesome and people should do it. Anyway, but if you want to make a real difference, I urge people to collaborate and to get out there and work together on things and to really seek, you know, in a collective impact approach with their technology ideas and solutions.

So, yeah. I thoroughly agree with solving your own problems or solving the problems of those around you, not solving the theoretical problems of someone half a world away. Your solution to, or my solution to water drilling in Africa is much, much, much less likely to be any use to anyone if I actually go to Africa and find out what they need

instead of just designing whatever looks neat in my backyard. Right. Or not understanding the environmental situation because you don't live there. Sure, you know, I can make the best drill ever, but it requires a nuclear power plant, which isn't that available. And there have been a lot of products like that where people, even well-intentioned and famous people, come up with grand ideas. I think Dean Kamen had a water filtration system like a decade ago, and it was hailed as this thing that was going to save Africa.

And I don't remember what the particular problem with it was, but you've never heard of it again because it didn't work. Required unobtainment. Right. It just required power or something that just wasn't ubiquitous in that situation. It just didn't work and you know when we you know when we do a lot of software and of course we're constantly having to think through low bandwidth low and no bandwidth situations um and you know as we work in 3d and we're beginning to do more work in india um in our global literacy work um we have to think about things like heat right um where manufacturing becomes or or tabletop

manufacturing can become very difficult in an environment where your climate is just too warm so i mean and wet don't forget the wet too and wet exactly exactly so absolutely there are a lot of environmental factors. There are a lot of cultural factors, community factors to take into account. And again, you know, I never want to discourage the tinkering in the garage. But, you know, impact comes from really looking outward as much as looking in. Yeah, really understanding the problem space. Exactly. really looking outward as much as looking in. Yeah, really understanding the problem space. Exactly. But some of it is about understanding the new technologies, the technologies that either are becoming cheap enough that they can become ubiquitous,

or the new, new technologies that may have a huge influence if we just could apply them properly. Do you have any technologies we should watch that are likely to lead to good benefit for lots of people? that these are the brightest, flashiest, shiniest things, but I think that they have great momentum and great capacity for changing lives. Machine learning is an easy one, right? That particularly in the disability space, where so much is about translating one stream of information into another stream of information.

So the visual into the audio, the, you know, or, or interpreting huge amounts of data that, you know, you otherwise couldn't. There's enormous potential there. And so that's an area that we are looking more and more into. We're also really interested in remote sensing and sort of implicit data gathering as well. Well, then you are going to need to get into embedded systems. I know. There's that, yes.

That may be two degrees removed from where we are now, but I don't think I agree with you. I mean, I think it's so far away. It's getting easier. It is really, yeah. It's getting easier, and it's something that we're watching very closely. So, you know, it's the other thing that I think is, let's see, what's the best way to put this?

What's interesting that we see, you know, we work with groups that are often grassroots groups. We're often working with NGOs and nonprofits that are, again, not the most technically savvy and or not the most well-funded. And what we find is, in many cases, the technology that is the most meaningful for them often already exists because they're already using it in some way and their their their constituents are already using it in some way um you know they don't want to have a secure data sharing service because they really do a lot of communication over Or that, you know, so increasingly, I think, because technology is so embedded in our lives already, it's not about the bright, new, shiny thing. It's really about better IT, right? It's about being able to. Oh, that's so depressing.

No, I mean, I don't mean to be depressing. And again, I'm not discounting the next great breakthroughs in technology. But when we're talking about groups that are not looking to, are really looking to do some fundamental things, in many cases, they just need that leg up, that they just need that small bit of additional technical guidance versus needing some fancy new communication system. Giant shiny object syndrome. Well, I think that

goes along with a lot of advancements, right? It's what can save time so that somebody's not constantly trying to fix something that's broken every day. You know, small advancements like, oh, I don't have to walk 600 feet to get water. You know, the computer equivalent of that. I don't have to spend two

hours a day making sure this is still running. Small things that free them up to do things that are more important. And that's the whole thing for us right now around 3D printing, right? You know, we think this whole tabletop manufacturing is going to, you know, it has such huge potential for the populations that we work with. But, you know, the last time I talked to a teacher about it, she said, oh, man, something else to go into the closet, right? Like, I'm not going to run a 3D printer. And that's a really good point. Babysitting technology has its own set of downsides. How do you decide when it's worth it? Or how do you decide when to convince somebody else that it's worth it? That's even harder. to really working closely with the communities and understanding what the tipping point might be.

So in the case of the 3D printers, I think there are certainly some schools and some teachers who are ready, willing, and able to step in and say, let's try our hand at this. But for all the others who still may want to take advantage of the possibility of having customized 3D objects in greasing a path for technology to help the most people. Have you had any failures? I know we don't usually like to talk about failures, but have you had any times at Benetech where you started a project and got a little further than just the analysis phase and then realized that you weren't helping you were getting in the way

we love failures in labs in labs yeah that's the best place for them exactly exactly i mean if we're not failing in some way then we're not trying hard enough. So, yeah, I mean, I think certainly we have had projects that didn't go the way we thought they would go. about a project we did that is trying to solve the problem of automated description of graphs in math. So, as you can imagine... I'm not even sure how to do that as a human. I know, I'm not even sure how to do that as a human. And then it goes up and to the left at a slope of about two, and then it stops and goes flat for a little while. And then it's not going to be easy. You got it.

We could have you do some templates. No, but the challenge was that we built it on top of software, which is great, which is essentially graphing calculator software for people with print disabilities. with Cartesian graphs in them that had to be retrofit to be accessible could automatically have these graphs converted into formats that would be accessible. that um let's just say actually you know it turned you know because it was based on a graphing calculator that the product itself um was pretty good if you if you were doing something from scratch if you were going to match up something that appeared in a book it was just it it didn't work well so so we learned a lot on that project. And I think the good news is we took maybe the UI

and the actual purpose of the product turned out to be a little bit sort of misguided maybe. Misguided is too strong a word. It didn't work as well as we had thought it would be, as we thought it would. But the underpinnings of it turned out to be extremely useful, which were these templatized translations of MathML into graphs into verbal description. And we're going to take that and adopt it in other ways. So, well, you know, we learned enough about how useful that component would be, but we scrapped everything else. When I wrote my technical book, I asked the publisher if I should do descriptions

for all of the images such that a blind person could understand. And the answer came back, no. And I have regretted that enough that we're going to fix it in the second edition. Wonderful. But it is really, really hard to describe a picture. I mean, if I'm talking about making a block diagram and I have two different ways to show I'm making a block diagram, because in the book I talk about different methodologies of that.

Well, I mean, the way I describe a system is the same in words, but visually is different, and that was my whole point, was that visually it's different, and if you can't see it... What advice do you have for software engineers and device makers for accessibility, visual accessibility, or people with visual issues? I agree with you. It's super hard. And in fact, for some time, we've had a program where we have volunteers come in to help describe images in textbooks and other books, including picture books, et cetera, in the Bookshare collection. And, you know, universally, people find it difficult. It's an art more than a science.

It's putting yourself into this space of trying to figure out what is important about an image so that you're not overwhelming the reader. It's very challenging. That said, because it's so challenging, one of the things that we have put together is a set of guidelines for people to help describe really common images, particularly common things like Venn diagrams and organizational charts and flows and just to give people a head start. And then we also have a free product called Poet, which you can find at poet.diagramcenter.org, which comes from our Diagram R&D Center. And Poet is a description tool for books, but it also has a training module where you can

go in and just get some help in terms of knowing when to describe and how to describe. So often, you know, one of the biggest challenges is not even how to describe something, but should I even describe it at all? And in many cases, it's not important to describe an image in a book or an article or what have you, just because it's eye candy. And just knowing that saves everybody a lot of time. But she also mentioned, and I don't know which publisher you might have been referring to, but if it was O'Reilly, I want to give a shout out to O'Reilly because they've actually partnered with us on math accessibility and are using a free tool that we developed based on MathJax called MathML Cloud to provide automatic transcription of math formulas in books. So because math, in many cases, because it's so typographically precise, is presented as an image. And that's a problem, right? You're running in a math book and you're converting it to both the typographically beautiful image

and a textual transcription. And O'Reilly is one of the first adopters of that, so I want to give them a shout-out. Cool. But what about software developers or device developers? Yeah, it's an interesting question because most application developers, probably this doesn't even enter their mind, right, when they're starting to develop an application.

So how do I make this accessible? You under such time pressure sometimes. And I know that Apple has made it quite easy on iOS for the iPhone and things to put accessibility hooks into your code. But that still requires thinking about it. Right. So that's a big topic. And so there are key resources for developers,

but mostly they're for web developers, right? So, you know, there are accessibility guidelines that come from W3C. There are accessibility checkers and guidelines that you can get whenever you're developing, you know, some front end for a website. Once you get into app development, it becomes more of a wild west. And you're absolutely right. Apple is a great place that has done an exemplary job creating accessibility core features in iOS that you can hook into. Android is getting better and better every day.

But the guidelines and best practices for those platforms are, I mean, they're not as clear. And so while there, you know, there are good resources out there, they are, they are pretty high level. But that said, I mean, I think as you're pointing out, for most software developers, getting sort of deep into the weeds of accessibility is not going to be possible anyway. And so there's just some key things to remember, right?

The first thing to remember is, you know, at any given time, create software that can be navigated, used, understood using multiple modalities. There is, so basically, can I, if I, can I do it, can I navigate and understand the software by sound alone? Can I do it by sight alone? Can I do it if I don't have access to a keyboard, but I only have access to, say, a push button or a flick switch?

Keeping in mind those different interfaces is important there are some basic core accessibility principles um that are called um poor p-o-u-r um so if you uh and then they include things on usability and lessons okay so we so we are having lots of Skype problems, and I know you have to get out of here. So let me ask you what your final thought is, and then Chris and I might talk a little bit more about accessibility. So do you have a final thought for us? Yeah, I don't think we've had a chance to really talk about

social entrepreneurship and social impact technology writ large. But I just wanted to say this is a great time to be in tech social entrepreneurship. I mean, the stars are aligning on a lot of different fronts. And one of the biggest gaps in terms of resources in the sector, yes, funding is always going to be an issue. But one of the biggest gaps is also talent. Especially in the Bay Area, it can be hard to compete with the likes of Google and Facebook and other organizations for engineers and technical talent. But there are huge opportunities right now to make a big difference by giving time in terms of volunteer time or, you know, making a career change into tech for good.

So I just wanted to sort of plant that seed that, you know, innovation in the tech for good sector is just exploding right now. And, you know, it's a great time to be exploring it. Yes, that's right. You don't have to be wondering if you should be battling your ethics and being forced to write VW-like emission scandals in the making. You can do things that are actually good all the time. Exactly. All right. Well, thank you so much for joining us.

My guest has been Anne Bowie, Vice President of Benetech Labs. Thank you so much. It was a pleasure. Okay. Now that Anne is off the line, which I'm sad about, but I understand how that works. I wanted to ask you, Chris, about some of the accessibility stuff in Apple. Have you seen the guidelines? I haven't seen the guidelines. I know the technical details of how to do some of the things in apps. Like what?

So there's voiceover. So you can, for every element on the screen, you can associate it with some spoken thing. So allows them as you scroll, there's system level accessibility things that, that you can turn on, uh, that allow you to have certain elements on the screen spoken to you and scroll through them in a predictable way. So you're not just randomly pushing areas on the screen. Um, well, there's some haptic feedback. I remember we watched a terrible TV show in which one of the characters, one of the heroes, was blind, and his interaction with his iPhone was awesome. He just held it, and it did what he wanted, and he couldn't see it.

So I know that that's possible. But for your app, a minimal set of things is giving a name to each element that makes sense this is the play button this is the play button this is the advance button this is you know stuff like that um but i wonder if people if most people writing apps even go to that minimum level of effort because i worry that there's no impetus beyond feel good, right? And yet, when I saw that on the TV, I was like, I should learn to do that. Then I can use my phone without looking at it. I can use my phone during meetings.

But it's one of those things where the accessibility features might help a lot more people if we were willing to learn them i mean it is a different it is something you have to learn to use in many cases and there's a lot of features beyond what i'm talking about that yeah that app developers can take advantage of and apple sort of quietly does that they don't make a big deal out of it i haven't seen a huge big deal out of it like like their presentation to WWDC or anything. They don't have it up there as, oh, we've added these new features as a line item. So I think drawing attention to it could help a little bit more. I think they're doing good work, but it's kind of, well, we're just doing great work that you can either use or not use. And I think Android, she mentioned Android,

but I think that's a longer, more difficult path because there's such fragmentation of just the devices, the operating system. It's very hard to write an app, you know, that might have to span three or four major versions of Android. You might not have some of those features available to you. And so I could see that as a, a well i'm just not going to bother it is well and i was really sad that we opted not to bother in my book because it was hard just because it is hard doesn't mean you shouldn't and i i don't want to say we we should all know somebody who has visual difficulties because that person will teach us so much about

making things easier but yeah you know cultivate that guy because or a girl because you know that's i i've lost it haven't i um it is worth talking to people well if you have no experience with it then you're not going to even think about it. And I think that's where the not thinking about it comes in. Because most people writing apps probably don't necessarily have that experience. And so it doesn't enter their mind.

It's not a bad thing. It's not that they're being bad people. It's that it's just not part of their experience. So they don't even go to look for how would I do this. And then at the end, when somebody says this is hard, they've already coded themselves into a corner such that it is very hard to change.

But if you know about it, maybe you can put it in. What would you think about the embedded space, though? I mean, because apps are... There's a way to do these things in apps. Do you think embedded devices make this easier or harder because some of them small devices don't have much of an interface they might be purely haptic or just have a couple of lights lights are a problem well this goes back to to colorblind when i i look at all the things that say

if it's flashing red do this if it's flashing green to that and i'm like well i know a guy who would just say it's flashing yeah but i mean i think there's an explosion of devices happening where there's all these small devices that are getting cheaper and more easy to produce the i hate to say the internet of things again, but I will, and I doubt most of the device manufacturers are thinking about accessibility of those, except for ones that are just automatically accessible to certain classes of disabilities, like the Amazon Alexa thing. Because it's voice activated. It's all voice. There's no other interface.

That works for blind people. Doesn't work so well for deaf people. think that that is the sort of thing that has already made a huge difference to accessibility being able to have an interface that is involved with moving things right instead of with letters or even with mice it's just i don't know it's really a lot better when we get into the haptic interfaces that we're sort of building for games, but those are going to be really useful for visualizing, except tactile. Is that a word? Tactilizing? No, that's not

a word. Making things so that you can touch them in order to understand how they're organized. 3D printing's great, but you can't print out everything. Sometimes you have bigger stuff. Right, I was just thinking when she was talking about if you have a book of math problems, and you want to print out everything. Sometimes you have bigger stuff. Right, I was just thinking when she was talking about if you have a book of math problems, you know, and you want to print out every graph, you can have a pile of X squared and X cubed. It's kind of an amusing thought.

But yeah, there's other ways to do that sort of thing where you have temporary, you know, she was mentioning two and two and a half d sort of things that um display technology it's sort of like braille except it's a more of a grid thing that you can touch but those get really expensive quick well i mean it's 3d pixels well and it's a low volume market yeah right and it's probably medical devices which boosts up you know in some sense which boosts up the price so yeah and yet i think it would be amazing to have something that i could touch yeah so maybe we need more art projects i hope the developers of all kinds at least at the beginning of a project think how could i make this more accessible and you you don't have to, this is like quality. You don't have to make it

the bestest, perfectest ever. You just have to make it a little bit better than you did last year. So, all right. Well, I'll end there. Thank you for continuing with me. Okay.

Now I will untie the chain that holds you to your desk I should send a thank you out to BJ for suggesting ANN and helping to set it up and also I should thank you listeners for listening hit the contact link on Embedded.fm if you'd like to say hello or email us show at embedded.fm

don't forget that Avid is looking for an embedded software engineer in Southern California. So that'll be in the show notes. And I'll talk to you next week. In the meantime, a final thought to leave you with from somebody who's left us. David Bowie said, don't you love the Oxford Dictionary? When I first read it, I thought it was a really, really long poem about everything. Embedded FM is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California.

If there are advertisements in the show, we did not put them there and do not receive any revenue from them. At this time, our sole sponsor remains Logical Elegance.