Soundbites of Graduate Student Research

So our lab works with people with spinal cord injury. Most people with spinal cord injury are incomplete, which means they have some volitional ability, they just need assistance. So the standard of care is physical therapy. The physical therapist can sense what they’re trying to do and provide assistance. An exoskeleton can do that, but never get tired. You can get in, you know, so many repetitions of your exercise, but it can’t tell what you’re doing. 

So when we have our patients with spinal cord injury walk with our exo and it kind of walks for them, they don’t really like it. It’s kind of constraining them. So we’re trying to make a model that reads muscle activity and predicts the future motion of the person and then [determines], “Okay, well, what do I do in response to that?” in terms of exoskeleton torque. 

So the muscle information that we use is ultrasound. Unlike EMG (electromyography), ultrasound can penetrate and read information from deep muscles. EMG, unless it’s invasive like a needle electrode, can only read the electrical activity of the surface. So we collected EMG and ultrasound data during two minutes of exoskeleton walking. We just had them walk with a very simple exoskeleton controller that’s trying to track a time-dependent walking trajectory. And we created these data matrices. So the data matrix is, we have a whole bunch of joint angles over time, exoskeleton motor currents over time, and then, depending on the model, ultrasound or EMG as well.

And you do some linear algebra. You get a linear model, and what a linear model means for us, working with like a multilink robot, is that it doesn’t crash your computer when you try to use it. So we found that if we use a model that only has the exoskeleton assistance torque and the exoskeleton angles it’s not terrible, but it does — the blue is the prediction — it starts to diverge over time from the actual angle. If we add the ultrasound echogenicity — which, what that means is the brightness of the signal, and it’s very bright in between muscles; there’s like a white layer that you can see — it increases the prediction accuracy by an average of 15 percent across participants, by 30 percent for our participants with spinal cord injury. EMG to my surprise did not improve prediction accuracy on average. There may be muscle crosstalk, like it’s getting information from other muscles, or it may be sensitive to placement. 

So, long story short, these models work pretty well. And now we want to plug them into what’s called model-based control, which is where, based on a model and the current information of the joint configuration and the muscle activity, it can make an optimal decision for assistance torque.

So this work, a lot of it’s done. I have some more data that I’m collecting but in terms of proof-of-concept, the A-mode sensor works with exoskeleton. We’ve got that. Now, I’m looking to close the loop, and finding the model was the hard part. I’m hoping, you know, fingers crossed, that now trying to plug that controller in and get the exoskeleton to move with the person, hopefully that goes smoothly. I mean, I should finish by August! 

So the VR fashion show basically is [one where you] put on the VR headset to look at the fashion show on a 360 degree virtual area. So, because I’m in the consumer behavioral area, I’m wondering, what is the consumers’ perception of VR fashion shows? Because we can see on the research, the VR market is very huge. From 2024, $67 billion jumped to $204 billion. So that’s huge. And brands, they always use the VR headsets to show they are different than the other brands. To touch more audience, right? To build a stronger brand engagement with the consumers. And for us, for the consumers, we want to have the experience, we want to know more, we want to have more things in our life. 

So, there’s a theory in academic areas called social presence, and the social presence means being with others. It’s very weird why they mention this in the VR area because it’s one person, one device. When you put it on, it’s actually you are alone. Why do they mention the social presence? Right? 

So I did a bit more research on the social presence theory. There are three dimensions. First one is the feeling of being with others. Second one is your emotion in this virtual environment, and the third one is your action there. So those are three original dimensions for the social presence. 

So, based on this, I have some interview questions. I want to know [how people are] feeling about this. So I did a two-focus-group study and each group has ten participants and each one has one VR headset. They watched the VR fashion show at the same time. And the VR fashion show will take totally like four minutes. And then we have one hour and a half to discuss how they feel about this fashion show. And those are my research methods, the data analysis methods. And then you can see the black parts are the original framework of social presence.

I got three new dimensions from this study to make this framework stronger. Yeah. The three new dimensions are escapism experience. That means when you put on the VR headsets, you will feel like you forgot all your problems in the real world. You forgot time, you forgot everything. You just enjoy immersive environment, right? And the second one is novelty and a VIP sensation. Like if you experience [the VR fashion show], you will know you are just jumping to the first row of the stage. You can see the model just right in front of you. So that’s a VIP spot. So most of the audience, they feel wonderful because they have this special event that they treat me like very special there. And then third one is social sharing and status signaling. Some of the audience, they mentioned that they want to have a screenshot and they want to put it on the social website. They want the other people [to be jealous of] them. To know they have this experience.

So those three dimensions, they are mentioned [as part of] the social presence but they are not included in the original framework. Yeah, so that’s huge. I build this. And then later the researchers, future researchers — not only in my research group, but also in the whole area — they can start to research based on this framework. It’s more stronger now. And then the industry, the brands can use this framework to know their consumer better. Right? 

And here at the right corner, here’s the universe that I actually built for the VR fashion show.  First off, for the first year of my Ph.D. program, I didn’t find that there’s enough categories for the VR fashion show, so I made a classification of VR. There are four types of VR fashion show now in the world. And then is this study: I just want to know [how the social presence shapes] consumer experience. And the third one is consumer perceived value. They viewed this VR fashion show, it must have some meanings for the consumer, right? I want to know how their values comes from when they watch this VR fashion show. 

And this one is my dissertation paper. For this one I actually recruit 117 participants. They actually put on the VR headsets, watch the VR fashion show and then help me to complete the survey. So currently I’m doing the data analysis and then I’m going to graduate in the summer.

And later I’m focusing on the Metaverse VR fashion show and AI fashion show. So currently you can see some AI generated fashion shows. They build the avatar but put the celebrities’ face on it. It’s very interesting. Like later on, maybe consumers like us, we can put our friend’s face on it, right? We can customize the fashion show. That’s very huge in the future. So later, I will focus on that. So that’s all my study and do you have any questions on that? 

Interviewer: No, I don’t think so. Thank you so much for walking me through it. 

Thank you. And do you want to experience it? 

Interviewer: I do, actually!

Think about this: 30% of Americans struggle to navigate government services. That’s nearly one in three people facing confusion, frustration and delays when doing basic things — renewing driver’s licenses, applying for benefits or figuring out public transit. When people struggle, it’s usually because something isn’t working, whether it’s outdated websites, unclear instructions or bureaucratic red tape. Governments are digitizing services but efficiency doesn’t always mean accessibility. 

That’s where service designers come in. They bridge the gap between agencies and the people they serve to make services more intuitive and inclusive. But service designers face growing challenges. Recent layoffs in digital transformation teams mean fewer specialists tackling these problems. Traditional research tools like personas and journey mapping are somewhat limited. Recruiting participants is difficult and insights often come from small and representative samples.

These methods struggle to keep up with major societal shifts like mass immigration, economic crises and changing tech literacy. The problem of non-inclusive government services will always be complex, but I wanted to look specifically at the role of design because the research phase is where customer experience gaps first emerge. If we improve how service designers understand and advocate for users, we can build public services that work for everyone — not just those who are tech savvy or persistent enough to navigate a broken system. 

My project explores how AI-powered adaptive personas and user simulations could help. And instead of relying on static user profiles, designers could work with dynamic, evolving personas that reflect real-world complexity. This would also allow them to test, refine and future-proof services faster, so government services don’t just meet today’s needs, but evolve alongside policy changes, demographics and technology.

Hi, I am Taynara, a Ph.D. student at the Department of Entomology and Plant Pathology. And I am a student at [Professor and Extension Specialist Dominic Reisig’s] lab and I work with stink bugs to understand the monitoring tools such as pyramid traps or pheromone traps to monitor them on soybeans. 

So stink bugs, if you’re not familiar with them, they are considered a pest of many fruits, veggies and row crops such as soybeans. And they estimate around $300 million in soybean costs and losses annually. Usually, they will show up after the flowering stage and they will feed on seeds and pods of soybeans. And it will not look any good, as you can see here. So, to manage stink bugs, we need to go out in the field and to scout using a beat cloth or a sweep net.

And if it [reaches] risk of economic loss, likely we will control them by spraying insecticide. However, monitoring tools such as pyramid traps could support our current scouting techniques. That’s why I had 68 fields in North Carolina and I placed two different pheromone lures in those fields that are commercially available — a single lure and a dual lure. And I had seven different commercially available pheromone traps that I tested in those fields. And I also scouted the middle of the fields using a sweep net. 

And then those are the four […] stink bug species that I found in soybeans: the brown, brown marmorated, green and southern green stink bugs. And I did all that because I had two goals: to determine which trap captured the most stink bugs throughout the season and also to determine which soybean growth stage stink bug population reached risk of economic loss. 

And now let’s go to my findings. So I found that the pyramid trap, also known as the black pyramid and the yellow pyramid, capture most stink bugs throughout the season. And when we checked the pheromone lure, the dual lure captured most stink bugs throughout the season. It attracted most species that I mentioned. And if we check the average number of stink bugs throughout the season, I found that there are more stink bugs that reached risk of economic loss from our four to our six soybean stage — full pod to full seed — when we have pods and seeds for them to feed on and cause damage. As nearly also one third of my fields were above risk of economic loss, out of my 68 fields. It’s a lot. 

And then for conclusion, I would recommend growers to use those pyramid traps with a dual lure. And we know that stink bugs numbers are higher from full pod to full seed, which means we should be considering scouting or putting those pyramid traps out there early in the season, maybe before flowering. Which is in here. 

And for implication, time-efficient sampling, as we can combine those pyramid traps with sweep net scouting. And also trying this in other crop systems because stink bugs damage not just soybeans, but fruits, veggies and other crops.

My degree work is based off of creating therapeutic foods for different populations. Therapeutic foods are foods that are designed for any specific medical condition or population. So this poster showcases six different types of therapeutic foods for two different overall populations — three for female adults and then three for older adults. 

For female adults, there’s one that was designed for the general population, for pregnant women and then children negative nine to zero months, and then for nursing women and children zero to six months old. For older adults, it was designed for the general population then for older adults in end-of-life care — not necessarily inpatient, but any adults with a terminal illness — and then individuals with dysphagia, which is like a difficulty swallowing. It was modified to kind of facilitate more comfortable swallowing and has a higher water content. All of these are processed with continuous-flow microwave heating, which ensures shelf stability, nutrient retention and safety.

The older adults are based off of purple flesh sweet potatoes. The female adult formulations are based off of orange flesh sweet potatoes. There will be a picture with this, as well? So that’s kind of the design (see poster). It’s been a little bit of a head scratcher, so I wanted to point it out. The right side is a cross-section of a purple sweet potato that I used during [research and development] and then the left side is a cross section of an orange [sweet potato]. Yeah, and there’s a lot of wonderful future work I would like to do, including clinical trials, sensory tests. But for now, these are the six prototypes.

I would love to do clinical trials and just generally know if people like the taste, too. I think they’re pretty decent, from my experience. We’re more worried about the kind of functionality and just, you know, hitting the mark on the nutritional profile and also shelf stability and safety. We’re going to work on, obviously, aesthetics later. I’ve gotten a lot of not so gentle comments around those. But yeah, for now, this is just kind of a prototype phase. But yes, on paper, definitely viable for a treatment plan. 

Five years ago, my advisor had a collaborator who came to us asking to develop a therapeutic food for children with severe malnutrition, globally. And so we worked the past several years developing that. And that was a similar technology, similar base to these, but wanted to expand them to different populations, different markets, in a human sense. I actually had the idea to work on some for endangered species as well. But the scale and reach for those kinds of population is not as big. So we decided to do therapeutics that would be better for an American market and reach more people. So we came up with these six.

So we all know that microplastic pollution is a growing concern nowadays. And that’s pushed scientists to explore bio-based textiles. Sodium alginate, extracted from brown seaweed, stands out for its biodegradability, and I’ll shorten it as SA for the rest of my presentation. However, pure SA fibers reported by previous studies are mechanically weak. And most studies try to fix this issue by adding either costly or non-biodegradable additives. And I asked, why not improve the process itself? So I ended up optimizing the process to make SA fibers from two angles — thermal needling and an organic solvent-based coagulation. 

So here’s how SA fibers are made before. People started with the SA solution in the water, extruded into a water-based calcium chloride bath where the calcium ion cross-linked the chemicals into a solid fiber. Normally those SA fibers still have to stay or age in the same bath for an additional hour to fully cross-link. But that traps water, which makes the fibers weak and makes them stick to each other. 

To address this issue, I replaced the aging step with a thermal treatment at 80, 120 and 160 degrees [Celsius], respectively. Result? Let’s move to the properties. 

This chart has the stiffness on the X, strength on the Y, while the bubble size represents stretchability of the fibers. It was found that as we increase our needling temperature from 80 to 160 degrees C, SA fibers become stronger and stiffer. At 160 degrees C, the fiber strength even increased by 149 percent, while stiffness increased by 86 percent, outperforming wool and approaching cotton. 

So, infrared spectroscopy here explains why, by showing right shift of the OH peak. This indicates that a stronger intermolecular interaction was formed as we increase our needling temperature. However, stickiness remained an issue for this method. So I replaced the water with methanol, a type of alcohol for coagulation. 

Why? According to Fick’s law, diffusion is driven by concentration gradient. With water-based coagulation, there’s no concentration gradient for the water, so water stays trapped inside the fiber. Methanol, however, creates a concentration gradient around the fiber so that the water will get a chance to diffuse out of the fiber.

I have the SEM images here on the fiber cross section, to further confirm this theory. As you can tell, when methanol is used, no residual water was found inside the fiber. And that really makes our fiber non-sticky. And furthermore, fibers made out of this way [are] softer and more stretchable — even more stretchable than all the other fibers on this chart.

And, in short, thermal needling boosts both strength and stiffness of SA fibers, while use of methanol removes stickiness and enhances stretchability of the fibers. Together, these two strategies create SA fibers with different but promising properties for two different potential applications in the textile industry.

And I have some fibers here you can feel free to touch.