SciTech Now: Episode 620

TRANSCRIPT

[ Theme music plays ] ♪♪

Coming up, technology tools to help victims of abuse.

And we're trying to work with the victims to figure out how to cut off those ties and how the abuse is occurring.

Improving critical thinking skills through technology.

One of the things we were hoping to do is to really break down critical thinking into meaningful and observable characteristics.

Creating a concussion sensor.

We keep it really simple -- If it's red, check your head.

Conserving water through technology.

We're essentially turning dumb water pipes into a smart water network.

It's all ahead.

Funding for this program is made possible by...

Hello. I'm Hari Sreenivasan.

Welcome to 'SciTech Now,' our weekly program bringing you the latest breakthroughs in science and technology and innovation.

Let's get started.

Victims of violent abuse are often partners, close relatives, or family members of their attackers.

It's known as intimate partner abuse, and there are many people working to protect and defend the victims.

Now researchers at Cornell Tech here in New York are partnering with New York City to develop technology that can help victims protect themselves.

Joining me now are two of the researchers working in this project at the Jacobs Technion-Cornell Institute at Cornell Tech.

Sam Havron is a computer science PhD student, and Diana Freed is a PhD student in information science.

Thanks for being with us.

So, what are the range of problems that people face when a relationship goes south these days?

So, the types of abuse that we're seeing are the physical abuse, the financial abuse, the emotional abuse and tied in -- and stalking behaviors, and all of that ties in with the technology abuse that we're seeing today.

And so people are coming to us with technology abuse because they are being tracked.

The person knows things about them that they didn't intentionally tell them.

And even when they escape the relationship, this person still has access to their private information.

So this is basically the kind of surveillance tape that we're all living in but used by someone who you no longer want to have a relationship with, they're using that on you.

Yeah, exactly.

And even after you've left this person, they're still -- You might have escaped them physically, but you haven't, in most cases, escaped them digitally.

There are a lot of traces still left, passwords that might have been kept the same and that you shared when you were still in a relationship.

And so what are the consequences of that?

What are the kind of abusive cases that you guys have seen when people are taking advantage of their former partners?

Well, people are looking to escape to sometimes discreet locations, sometimes to shelters, and they're in relationships that they have chosen to end, and they can't escape because the person might still be logged into their device.

There might be shared children involved where the abuser has gifted the child a phone, and that phone is going back and forth.

And that is also being used to surveil the other partner.

And so we're seeing a lot of cases where there is still information being shared, social media accounts being accessed, and we're trying to work with the victims to figure out how to cut off those ties and how the abuse is occurring.

So, since this isn't this -- Let's say it doesn't include marriage or in a court of law, do the technology platforms get involved at all?

If somebody calls and complains to -- whether it's to Google or Facebook or wherever and say, 'Hey, I'm in this situation where I'm now being stalked and my information's in the wrong hands'?

Yeah. I mean, for the most part, these platforms will have ways where you can report someone who's logging in, 'And it's not me.'

And typically, companies will say, 'Okay.

You can try to change your password and maybe that will help you.'

And that's usually the end of what they offer.

Yeah. So, YouTube built a piece of software.

Let's take a look at what that does.

OK. So we've plugged in a phone that might have -- This could be a standard victim's phone.

Right?

And you've decided to create what a tool does.

How does it look into this phone?

Yeah. So this tool has programmatic access to Android and iOS devices.

And what it does is, it looks for the kinds of applications that might be used to spy or track or surveil people.

And so...

It's trying to scan that.

Yeah. So how it'll work is that it will talk to the phone and ask it for what applications it has installed on it.

It will also check to see if the phone is jailbroken or rooted, which can indicate some more serious things that are going on.

So all these are apps that, in this column here, that have ways for someone to surveil that phone.

So this is... That's a lot of different apps.

I know this is a particular phone that you've used for testing, but there's lots of different types of applications where another person can spy on you.

And this piece of software helps identify all of the different pieces of software.

Now, the choice, of course, is for this person, in their phone, they could just go ahead and delete that.

But that actually sometimes becomes a bigger problem because then the person who put it on there knows, and it gets complicated.

So how do you deal with that?

Which is kind of a tense situation.

Yeah. So, when we're running, our consultations with people who come to meet with us, our whole model has in place, it's a referral model.

So people who come to us have existing advocates for them -- case managers, lawyers, et cetera.

And so before we make any serious change, like deleting an application or changing a password or something to that effect, they need to have a conversation with their advocate to do what is called proper safety planning so they can assess the risk of what's going to happen if you delete that spyware application.

Will your abuser know that you did that?

How will they react once they find out that you did that?

Is that a safe move for you to do?

And how do you get the software out, and where is it going now?

I know this is something that you've been working on in New York City.

Yes.

But what are you learning, and where's it going to spread?

So, we're learning the different types of -- types of software that are being installed on victims' phones, and what we're doing is we're open-sourcing all of the materials that we develop to run the clinic, and we're making it available to other advocates working on these types of cases.

So this might -- Right now, in New York, it's available where?

Our work currently is occurring at the Family Justice Centers, which is run by the mayor's office to end domestic and gender-based violence.

So we currently run our clinics in the five boroughs within the confines of the Family Justice Centers.

And we are now making all of our research publicly available.

So all of our tools to other clinics and we will help other advocates work with these tools to scale their own work.

So, a domestic-violence shelter in some other part of the country, the code on how this is built is freely available, and they can build it themselves.

And if they got questions, they can ask you about that.

Absolutely.

So what's the end goal here?

Your best-case scenario, five years out from now, how is this software being used?

So, I guess the goal is that we want to have software that is evolving so that if there are spyware applications, for instance, that are aware of this tool and trying to figure out if it's being used on the device, like it can become an arms race.

And so actually, part of the reason why we developed this as a desktop application is that typical anti-virus or anti-spyware applications, they're installed on a phone, then some of the spyware applications could report, 'This app is installed.'

So we made sure that, as a safety measure, there's no trace of this tool being used on the phone.

Okay.

So, several years from now -- Well, first of all, we hope that there's not a need for this.

But you know that technology evolves and abusive behavior will always continue.

So how do we -- How do you hope that this changes that landscape?

Well, at the very least, we hope that this is something that can help empower victims so that they are aware of tools or applications that might have been installed on their device that they didn't know about, and then they can start to make a decision for themselves as to what they should do next with that.

And right now, there really aren't tools like this.

There's really no resources for victims to get help with technology abuse in the way that these clinics are doing.

And so the hope is, like, five years from now or however long, they're going to start to see a change where hopefully anyone who is suffering from technology aspects of abuse from IPVs, they're going to have resources available to them.

All right. Sam Havron, Diana Freed, thanks so much.

Thank you.

Thanks.

Data science, along with machine learning, the two of them together, represent 5 of the top 15 growing jobs in America today.

So if you look at that list of the top 15, five of them are data and machine-learning related.

The second thing is that they are always accompanied by severe skill gaps.

So, all the jobs that exist in San Francisco, including the ones that haven't been filled yet, there are more than 38,000 jobs which need to be filled more than there are people who have the skills to fill them.

So if you're learning data science, however you're actually learning it, whether you're learning it here or whether you're learning it through some other mechanism or through experience, there is a gigantic skill gap waiting for you out there.

We're going to have a TV audience.

A lot of people might be novices to this topic.

Yes.

A lot of people might conflate big data with data analytics.

It's different, right?

How?

So, big data is the foundation on which it's all built.

This ability to not only see but also store all the data necessary.

So, there was actually a new company which went pub-- not went public, but it went out of stealth today, which talked about storing exabytes of data -- as a service.

And that's exactly what people need to do right now.

Big data means, 'I want to store everything I know about my supply chain or everything I know about how our users use our products all together so I can look at it in real time, and I can look at it historically over the entire life of the company.

That's big data.

So, way bigger than any data that anyone had ever managed before.

Data analytics.

And data analytics is the process of analyzing that data and drawing conclusions from it.

And machine learning is the process of building algorithms on top of that data to either optimize for particular kinds of outcomes or predict behavior.

It sounds like you need a computer science degree to have that kind of job.

Do you have to have a computer science --

You do not.

So, in the early days of LinkedIn, we were having a hard time hiring data scientists, the same as today.

And data scientists in 2005 and 2006 almost entirely came from the world of the hard sciences.

Physics, chemistry...

Biology and so forth.

And that is still the case today.

So, if you are in one of those fields, you have an opportunity to basically convert and become a data scientist.

You've already learned how to think about it properly.

You just need to learn the tools.

I think that would be a big help when it comes to increasing the number of women in data analytics, right?

Because when you look at computer science degrees, they're almost all men.

Yes.

But when you look at the hard sciences, there's a much more even breakdown, isn't there?

It really is, and this is -- this is a major topic.

So, right now, there are four men working in data science for every one woman -- data science and machine learning.

And that happens to be about what the proportions are in the world of computer science.

There's a real gap to be filled there.

The Natural History Museum of Utah is using technology to help students improve their critical-thinking skills.

Here's the story.

♪♪

Research Quests are a series of online investigations that allow students to tackle some of the same juicy questions our research scientists work on studying the same objects they use and using the same data they use to explore those questions.

Research Quest got started through a partnership with some of our funding partners who had a shared interest in ways that we could leverage digital technology to help support critical thinking in learners.

One of the things that we were especially interested in are the ways that digital technologies could help us expand our reach.

We've been working with public schools for 50 years through face-to-face programs that have been wildly successful in our state, but we never have been able to meet the demand for our programs.

So we hope that Research Quest can be a way for us to reach a larger audience.

One of the things that's been interesting about Research Quest is, we've worked every step of the way with learning scientists to study how students are learning with the resources we provide.

That has allowed us to work through iterative design cycles to get to our final investigations and then to study the learning that's happening.

We can ensure that these investigations aren't just engaging but they're also moving the needle on advancing critical-thinking skills.

So, Research Quest is really designed to enhance students critical thinking.

We like to say critical thinking, but it's really hard to define sometimes.

So one of the things we were hoping to do is to really break down critical thinking into meaningful and observable characteristics, to look at how those were occurring as students worked together during the Research Quest, and then to see how those critical thinking outcomes could be observed after the Research Quest, both in what students expressed about how to conduct research, how they understood evidence, and how they approached questions and trying to answer them with evidence.

My role in Research Quest has been to consult on the learning aspects to really try and design Research Quest to support deep and meaningful learning processes, especially those surrounding critical thinking, as well as to evaluate whether those processes are occurring in the classroom and the outcomes associated with critical thinking.

In the first sets of Research Quest, when we first started out, what we found was some of the maybe easier aspects of critical thinking were fairly natural for students.

They were really engaged with the museum objects, so they asked a lot of questions.

They developed hypotheses to test, but they didn't always engage in the more difficult aspects of critical thinking, and those are things like flexible thinking, keeping your mind open until you have sufficient evidence so you're not making up your mind too fast.

And also to really think more carefully with evidence and the quality and the balance of evidence, instead of finding maybe a piece or two of evidence and then deciding that they must be right.

We designed Research Quest over time to try and enhance those more deep and difficult aspects of critical thinking.

We look at the process of what students were engaged in in the classroom and really try and think about 'How can we create materials, prompts, situations in Research Quest that really helps students engage in those more difficult aspects and to hold back and learn to use evidence in different ways.

One of our investigations is situated at the Cleveland-Lloyd Dinosaur Quarry here in Utah.

And one of the basic phenomena that students explore in this investigation is trying to figure out what dinosaur particular bones have come from that have come out of that quarry.

Compare your earlier observations with the identified fossils.

Are the features the same or different?

For example, do a few species jaws have teeth similar in shape to the ones in our mystery jar?

That's helpful weak evidence that can start to point you in the right direction.

The process of developing Research Quest was very rich and quite iterative.

We did sketches and low-fidelity models.

We did videos.

We also did a whole bunch of video and audio analyses from our research classrooms.

So we engaged in a really detailed analysis of students' conversations and looking at how they responded both to materials and to the support in Research Quest, then changing and revising those materials in order to stimulate the kinds of conversation that we know would be beneficial for students.

This is a brand-new program for us, so it's been a an exciting and steep learning curve.

Unlike our other programs, it requires us to work with a large variety of partners.

We have scientific experts who are informing the science behind the investigations.

We've got media specialists who are helping us produce videos.

We have interactives developers who are helping us visualize the data.

We have web developers who are helping us get it right so that students have an interface that's usable.

We're hopeful along the way that we may be able to identify a framework that will make producing future Research Quests easier, more efficient, in a way that we can share with other museums so that they, too, could build these sorts of resources.

One of the most rewarding parts of Research Quest, besides the fact that I get to work with a wonderful team of experts, are the teachers.

Teachers are under-resourced and overworked, and they have expressed so much enthusiasm for having a turnkey resource that's very easy for them to use and highly engaging for their students.

If you're a teacher, you can access Research Quest by going to the website, creating a free account, and what you'll find is access to those investigations for your students.

You'll have teacher support materials, you'll have student notebooks, and you'll also have assessments for measuring the learning that's happening in your classroom.

There are lots of learning tools that were designed to be exciting and fun for students.

But I think the process of developing Research Quest where we use data at every step of the way to really hone in on effective learning processes.

How does each component of Research Quest help students cognitively engage in the kinds of thinking that are important and to communicate that with others?

That process level of development is really unique, but very important.

I think we could use more of that in educational technology.

♪♪ ♪♪

A young entrepreneur is disrupting the way we detect head injuries in sports.

After her own head injury, she rolled up her sleeves and got to work, creating a purely mechanical head impact sensor.

PBS NewsHour Student Reporting Labs takes us inside the factory to learn more.

I was in a game, got hurt, had no idea that I was hurt, and I was looking into, like, a head-impact sensor that could tell me if I had a -- you know, potentially have a concussion or not.

And the thing was, there was a mouthguard on the market.

It was the only product at the time.

Cost $299.

And I was like, 'I go through three mouth guards a season, like, losing them, chewing through them.

There's no way I could afford this.'

And I was like, 'There has to be a different way to detect head impacts and concussions.'

My name's Jessie, and I'm the CEO and founder of Tozuda.

At Tozuda, we design, sell, and manufacture head-impact sensors.

We keep it really simple -- if it's red, check your head.

It turns red.

You know, this is the time to stop playing.

This is a time to go to the doctor and get checked out.

So, we have two balls that are held in place by a spring, and there's dye in the end caps of the chambers, on the right and left side of the device.

So those balls will move independently like your brain does and will either compress the spring out of place or rotate the spring out of place.

I was never in the financial position to, like, just throw money at this.

so I was like, 'You know what?

What if I work -- I work for three years?'

So I actually worked for my family's business for three years.

And then, like, the nights and weekends, I was always trying to, like, prototype and iterate and try to figure out a way to detect forces and impacts without electronics.

Concussions are an invisible injury.

You know, if someone would have just pulled me from that game when it happened, I wouldn't have suffered.

And I'm hoping that this product could at least get people to proper recovery and to the doctor's faster.

♪♪ ♪♪

The average American uses more than 80 gallons of water every day.

But smart technology can help us all conserve water.

One company has developed technology to help track water usage and even give tips on how to use less.

Here's how.

Using technology, we're essentially turning dumb water pipes into a smart water network.

'Pani' actually means 'water' in Hindi, and we're building smart home water tech products and platform that measure, monitor, and eventually recycle water almost in that exact order.

So I started the company last year after a volunteer trip to Nepal building water wells for this tiny remote village.

And, you know, until you spend some time in a place where there's no infrastructure for any electricity or water, you realize how much you take for granted the things you have in everyday life.

I came back, and I started Pani.

Our first product is a smart -- smart home water monitor that you can install it on your shower head, on your toilet, anywhere you use water inside the home.

And when the water flows through it, it turns a magnetic turbine, which wakes up our Wi-Fi enabled smart home device, and we start to send data into the cloud.

And we have a free companion mobile app that shows you how you're doing in real time by individual fixtures and appliances, how you're using water.

It takes roughly anywhere from 1,800 to 2,200 counts to mean a gallon.

And we've done a lot of testing and in-house validation and calibration of these counts to determine that there are slight differences in those revolutions, depending on whether or not it's on a toilet or if it's on a shower, because a shower has a bigger pipe, so the water has a little bit different flow than if it was on a toilet, which has a narrower type of tube that goes to that device.

And so we've done a lot of calibrations to get as accurate as we possibly can with the number of counts that equals a particular gallon.

So we can give you as accurate an information as we can.

What we like about the Pani is its ability to get real granular level detail on water usage.

It works, you know, quite efficiently and delivering that kind of that ongoing coaching.

So once it's installed, it may give you tips about, 'Hey, did you know that it takes, you know, .04 gallons to grow one strawberry?

The next day, it may tell you, you know, 'You can, you know, before the shower heats up, you can take that one gallon or two gallons of water, use that to water your plants.

And that's a relatively low cost.

So a homeowner can start with one of these units and install it in one shower.

They can then install it on a few more fixtures in the home.

The water that you're saving, the Pani Systems company will actually donate on your behalf to these global water projects.

So, in our case, since we put this product in, I think we've added something like 236 gallons to that project.

It allows you to save water at home.

But on the other hand, you're helping someone in another country get access to maybe clean water, you know, for the first time.

[ Toilet flushes ]

You know, between your showers and your toilets, that's more than half of all the water use.

And once you understand how much water you use, then you make a conscious and a mindful effort to use a little bit less for the benefit of not just you and your home and your water bill but for all of mankind.

And that wraps it up for this time.

For more on science, technology, and innovation, visit our website, check us out on Facebook and Instagram, and join the conversation on Twitter.

You can also subscribe to our YouTube channel.

Until then, I'm Hari Sreenivasan.

Thanks for watching.

Funding for this program is made possible by... [ Theme music plays ] ♪♪ ♪♪ ♪♪ ♪♪ ♪♪