Experts Matter. Find Yours.
Connect for media, speaking, professional opportunities & more.
This article is republished from The Conversation under a Creative Commons license. Read the original article here. The last few weeks of summer, heading into Labor Day weekend, can sometimes mean vacations and driving more miles on the road for all people, including teens. Traffic crashes are the No. 1 cause of death for teens, and the crash rate for teen drivers is disproportionately higher than the share of licensed teen drivers. In addition to this grim statistic, summer is the riskiest time for teen drivers. The 100 deadliest days represent the period from Memorial Day to Labor Day when the number of fatal crashes involving teen drivers dramatically increases. A third of each year’s teen driver crashes occur during the summer. We are scholars who research transportation safety and teen driver behavior. Our expertise helps us understand that these 100 days are not just a statistical fluke – they reflect a dangerous intersection of factors such as inexperience and a propensity to take risks. What makes summer different? Regardless of the season, some teen drivers engage in risky behaviors that increase their likelihood of a fatal crash, such as getting distracted, driving with friends in the vehicle, driving under the influence, not wearing seat belts and a lack of hazard awareness. Teens also have more free time in the summer, since most aren’t in school. Combined with the longer days and better weather, teens drive more over the summer. More time on the road means more risk, especially for inexperienced drivers. Teens may also be more likely to drive after dark during the summer, in comparison to more experienced drivers. But nighttime driving is also when visibility is reduced and crash risks are higher, particularly for teens who haven’t fully developed the skills necessary for night driving. This increased exposure, in addition to teens’ general risky driving tendencies, contributes to the 100 deadliest days for teen drivers. The increased crash risk for teens over the summer isn’t equally distributed either. Crashes with teen drivers that lead to serious injuries are more likely to occur with male drivers, in rural areas, for those of lower socioeconomic status and for those with disorders, such as attention deficit and hyperactivity disorder. Teaching young drivers Driver’s education programs are the formal method to teach teen drivers the rules of the road. In driver’s education programs, teens receive information about driver and road safety though classroom and behind-the-wheel instruction in preparation for the licensing exam. Some states require teens to complete a driver’s education course if they want to receive a license under the age of 18. Of teens who have a license, nearly 80% of them have gone through some form of driver’s education. Though driver’s education programs can be helpful, their effects are not equally felt. In some states, teens and their guardians must pay out of pocket for driver’s education courses to obtain a license. This makes driver’s education and, as a consequence, obtaining a driver’s license inequitable. There are also driving school deserts – areas where the poverty rate is 20% or above and there are no behind-the-wheel driver education courses within a 10- to 15-minute drive. This makes driver education courses inaccessible. Many of these driving school deserts happen to be in areas with high populations of minorities. Over 20 years ago, graduated driver licensing was introduced to reduce teen crash rates. This is a phased licensing system wherein teen drivers are restricted in terms of when, where and with whom they can drive until they turn 18. Such a system allows teens to gradually learn and gain experience with driving over time. Graduated driver licensing has been implemented in all 50 states, and it has been shown to reduce teen driver crash rates. However, its effectiveness is limited to those who participate in the system. A large number of teens are unlicensed and are of low socioeconomic status. Many of these unlicensed teens forgo the entire process and remain unlicensed but still drive, well into their 20s when the graduated driver licensing restrictions are lifted. Making summer safer There are two things people can do to turn the 100 deadliest days into the 100 safest days. First, it is important that communities offer free supplementary training programs for teen drivers, because becoming a safe and responsible teen driver shouldn’t be limited to those with resources. As one example, in collaboration with industry partners, we have developed a program called Risk-ATTEND. It is a free, online, evidence-based program that teaches teen drivers how to anticipate risks while driving. Our research has shown that programs such as these can improve teen driving skills and may be especially effective for teen drivers in high-poverty areas. Second, our research has shown that parents and guardians still play an important role in influencing teen driver behavior. Studies show that teens mirror the behaviors they observe: If they see adults text and drive, they’re more likely to do the same. Once teenagers become old enough to drive, it is also important to establish rules and guidelines about expectations to establish clarity and accountability. Written agreements or checklists can address high-risk conditions such as nighttime driving, driving with other young passengers, phone use and adherence to speed limits. Systems to help monitor and enforce rules have been shown to be effective in improving teen driver behavior. One such program is Checkpoints, which is a Connecticut-based program in which families agree to limit teen driving during high-risk conditions. Teens face consequences for violating these limits, such as a temporary loss of driving privileges. However, the limits are gradually lifted as they gain driving experience. More than rules matter Ultimately, preventing crashes in the summer and beyond extends beyond mere adherence to regulations. Avoiding them fundamentally hinges on cultivating a robust safety culture that emphasizes a collective commitment to risk reduction and continuous improvement in driving practices. For teens, the summer months present unique challenges and opportunities. Drawing on best practices, such as training programs, teens can build essential skills in varied conditions before gaining full, unsupervised privileges.
Five storms predicted in the Gulf this season, says CoMET Lab
A composite of 500-mb temperatures from the seasonal weather model simulations used to generate the forecast. (The green outline shows the area averaged across for the Gulf of America.) The "-4.85C" is the temperature over the Gulf that ultimately fed into the equation to yield the prediction of 5 named storms. This hurricane season, five named storms are predicted for the Gulf, according to LSU’s Coastal Meteorology, or CoMET Lab, which issues a Gulf of America-specific forecast. Five represents a slight uptick in storm activity, said Paul Miller, the head of the CoMET Lab. Miller is an associate professor in the Department of Oceanography & Coastal Sciences, or DOCS. The seasonal average for the area is 3.7 storms. This forecast in elevated hurricane activity is in alignment with predictions for the Atlantic Ocean as a whole – NOAA has called for between 13 and 19 storms in the basin between the beginning of June and the end of November. CoMET Lab’s forecast works a little differently than the Atlantic-based predictions. Those models generally focus on ocean temperatures, among other factors. CoMET’s forecast, which was initially developed with Geography and Anthropology Professor Jill Trepanier, instead focuses on average atmospheric temperatures around the Gulf, approximately six kilometers, or four miles, above the ground. Miller said he and Trepanier considered a range of potential predictive factors when they first developed the forecast in 2021, before settling on atmospheric temperature. “What our data shows is that the warmer the temperature is, the fewer storms there should be. The colder the temperature is, the more storms,” Miller said. This link to cooler temperatures at high atmospheric altitudes may appear counterintuitive, but, he said, “think of a hot air balloon. It only works if it’s hotter than the air around it… If you go four miles above the ground, that tells you if you have a reservoir of potential buoyancy. If the air can make it to that altitude, it’s got a better chance of continuing to rise and forming [a storm].” Normally CoMET’s forecast utilizes temperatures throughout the month of May, but due to recent changes in data available from NOAA, Miller noted that this year’s forecast is based on readings from the last two weeks of the month. Article originally posted here.
LSU Lab Helps Louisiana Prepare for Hurricanes, Drought, even Saharan Dust
Any hurricane that forms in the Gulf of America is a direct threat to Louisiana and its neighboring states. But most seasonal forecasts focus on the entire North Atlantic Basin, including areas where storms may never come close to any land, much less the U.S. Gulf Coast. A Gulf-specific forecast developed at LSU’s Coastal Meteorology (COMET) Lab addresses that issue by providing storm information specifically geared toward the Gulf region. The lab is run by Paul Miller, an associate professor at LSU who founded it in 2019. “We decided that a Gulf-specific forecast could help state officials and Gulf-area residents better understand how active the upcoming season might be in their part of the Atlantic,” said Miller, who teaches in the Department of Oceanography & Coastal Sciences in the College of the Coast & Environment. Miller said the LSU-Velocity Risk Gulf Hurricane Outlook is one example of how the COMET Lab delivers real benefits to Louisiana. “A lot of forecasts tend to align with each other each year. But ours works a lot differently than some of the other forecasts that are geared towards larger areas of the ocean,” he said. “We’re not the world’s leading hurricane research lab—and we don’t try to be. Instead, we prioritize meeting the largest research gaps relevant to Louisiana residents, spanning a wide variety of weather hazards. “We want to make sure Louisianans experience a clear return-on-investment from our lab’s activities.” Ways the lab supports the state also include: Helping forecast storm surge in real time and informing the decision on when to close flood barriers before a tropical system hits land. Developing rainfall models to support flood prevention efforts. Studying weather patterns that cause drought and low coastal water levels that can lead to marsh loss during dry spells on the Gulf Coast. Saharan Dust Research One area of study that incorporates both air quality and tropical weather is the lab’s research on Saharan dust clouds, which are blown across the Atlantic in an air mass called the Saharan Air Layer, or SAL. “So, this sort of far-off distant concept of Saharan dust is actually something that is kind of important to folks here in Louisiana,” Miller said. This dust can shut down thunderstorm activity in the U.S. Caribbean territories, a key area of research in the COMET Lab, and cause respiratory problems when it reaches the Southeast U.S. The SAL can also suppress hurricane activity in the Atlantic. “Our lab just launched a new project with the Office of Naval Research to determine how dust-dimmed sunlight can affect ocean temperatures in the Atlantic’s most active hurricane breeding grounds,” Miller said. Read the full story here.
LSU’s Jill Trepanier Educating K-12 Louisiana Students About the Environment
What began in 2018 as a single rooftop weather station on LSU’s campus as a tool to help freshmen connect to the science happening around them, has grown into an educational network in the southern part of the state, connecting K-12 students with the sky through real-time data, interactive technology, and hands-on learning. Trepanier, a professor and department chair in LSU’s Department of Geography & Anthropology, leads a project that now includes 10 weather stations installed at or near K–12 schools from Lake Charles to Grand Isle. “The environment is harsh in Louisiana. Beautiful, but harsh,” Trepanier said. “The more students know about it, the better they can protect themselves and their families. For me, that’s what it is all about.” The project all started to help college students in Trepanier’s meteorology and physical geography classes connect more deeply with the material by using weather data collected from the air around them. “These were 400 freshmen every semester who were not geography majors, so they didn't really love the science of the atmosphere. But they were able to connect with the information because they could see the data on an app on their phone as they were living in it.” Installed across South Louisiana, each weather station is solar-powered and connected to a console that uploads data to an online web platform and displays it on a dashboard. Then an app shows the local students the current conditions and records for the day. “When we look at data from the community, it might be many miles from where you are. And most people live within a few miles or less of their schools. It allows them a close-up view of what is happening, instead of relying on something miles away,” she said. Teachers can use the data with certain lessons or during a passing storm. But the available data also educates them on things like solar radiation, “It also helps aid things like seasonality and our relationship with the sun. It extends well beyond just rain.” The material is also aligned with the Louisiana Student Science Standards for environmental and Earth sciences. “By allowing students to compare real data across space and time, it helps them to understand how systems are connected. And most of these science standards have them focusing on system theory, in one way or another,” Trepanier said. Read the full story here.
Professor Gina Rippon signs a copy of The Lost Girls of Autism for talk attendee Dr Georgie Agar Professor Gina Rippon’s new book, The Lost Girls of Autism, investigates why autism was thought to be a male condition for so long She gave a public talk at Aston University on 6 May 2025 exploring the central themes of the book Women and girls with autism have long been overlooked as they are better at masking and camouflaging so ‘fail’ standard tests. Autism in women and girls has been overlooked for decades, and Gina Rippon, professor emeritus of cognitive neuroimaging at Aston University Institute of Health and Neurodevelopment (IHN), has given a talk about her new book on the topic at Aston University. The book, The Lost Girls of Autism, was released on 3 April 2025, coinciding with Autism Acceptance Month, with the subtitle ‘How Science Failed Autistic Women and the New Research that’s Changing the Story’. Autism is characterised by a number of now well-known traits, including social awkwardness, extreme obsessions, and unusual movements and coping mechanisms known as ‘stimming’. It was (allegedly) first described in the 1940s separately by Austrian psychiatrists Leo Kanner and Hans Asperger. Originally identified as a rare developmental condition, since the 1980s, there has been an 800% increase in diagnoses, leading to concerns about an ‘autism epidemic’. There is a strong and enduring belief that it is a condition much more prevalent in males. Professor Rippon described her research as “looking at how brains get to be different and what that means for the owners of those brains”. This includes looking at the functions of different areas of the brain using scanners. During research into a number of brain conditions and diseases with obvious differences between the sexes, including how the disease progresses, such as Alzheimer’s in women, or prevalence in one particular sex, such as Parkinson’s in men, Professor Rippon also became interested in autism, also assumed to be largely a condition in males. However, during a research review, she found that many autism studies made no reference to sex differences. Amalgamated data from autism studies found that 80% of participants were male, and 25% of testing centres only tested males with autism. By only looking at males, Professor Rippon explained, the notion that autism is a male disorder became self-fulfilling. This does not just refer to scientific research. Even now, boys are ten times more likely to be referred for assessment for autism and twice as likely to be diagnosed than girls, even when they have exactly the same traits. 80% of autistic females have received multiple wrong diagnoses, including borderline personality disorder, social anxiety or obsessive-compulsive disorder (OCD). But why? The reason is the unchallenged belief that ‘autism is a “boy” thing’ causing a male spotlight problem in all aspects of the autism story. It could also be that females with autism express the condition differently. Professor Rippon said: “This took me back to [my previous book] The Gendered Brain when I was looking at the very clear view of what males should be like and what females should be like. If you look at the autistic population you have this clear idea that males are like this, but females, er, not so much? Females have poor social skills, but not as poor, or obsessive interests, but not as obsessive, so the trouble with females, is that they are not autistic enough.” The gold standard tests for autism are the Autism Diagnostic Observation Schedule (ADOS) and the Autism Diagnostic Interview (ADI) tests. Professor Rippon believes these are heavily biased towards how the condition manifests itself in males, such as social awkwardness or extreme obsessions. For example, parents may well be asked if their son has an unusual interest in weather patterns or train timetables, but they are not asked if their daughter has an unusual interest in Barbie dolls, because dolls are seen as socially acceptable. Research has shown that females with autism are more likely to ‘camouflage’ their symptoms, watching how ‘normal people’ behave, even practising social interactions, so they appear more normal. They are also more likely to ‘mask’ symptoms behind a persona, such as the ‘class clown’ or ‘star athlete’, in an effort to fit in. Autistic females describe this behaviour as a ‘survival strategy’ to avoid being spotted as different. It is also the case that girls are more likely to have sensory processing problems, such as aversion to strong smells, which can be enough to affect their day-to-day lives. This has only recently been added to the diagnostic criteria for autism. If the camouflaging or masking collapses, rates of other conditions such as disordered eating or anorexia, self-harm and gender dysphoria are disproportionately high, and it is these which will become identified as the underlying difficulty, rather than autism itself. Professor Rippon said: “The next stage should be asking why this group of individuals persists in hiding their autism, especially when autism has been defined as a lack of interest in social connection. There’s what I call the ‘born to be mild’ effect, where little girls are trained to socialise more, to behave, not to make a fuss, if you feel uncomfortable, don’t tell anyone else about it. There’s a lovely comment from one late-diagnosed female who rues the fact that she was so well behaved and wishes that she had just burned more cars so that someone would have spotted her carefully camouflaged distress!” The final slide in the presentation covered what Professor Rippon called “an ironic footnote”. While Leo Kanner and Hans Asperger are described as the fathers of autism, writing in the 1940s, it was in fact a Soviet female psychiatrist, Grunya Sukhareva, writing in the 1920s, who first described autism, even clearly examining the differences in the condition between boys and girls. Why her research was ignored for so long is unclear, but the male spotlight problem may well have been avoided. For more information about The Lost Girls of Autism, visit https://www.panmacmillan.com/authors/gina-rippon/the-lost-girls-of-autism/9781035011629.
NASA Asks Researchers to Help Define Trustworthiness in Autonomous Systems
A Florida Tech-led group of researchers was selected to help NASA solve challenges in aviation through its prestigious University Leadership Initiative (ULI) program. Over the next three years, associate professor of computer science and software engineering Siddhartha Bhattacharyya and professor of aviation human factors Meredith Carroll will work to understand the vital role of trust in autonomy. Their project, “Trustworthy Resilient Autonomous Agents for Safe City Transportation in the Evolving New Decade” (TRANSCEND), aims to establish a common framework for engineers and human operators to determine the trustworthiness of machine-learning-enabled autonomous aviation safety systems. Autonomous systems are those that can perform independent tasks without requiring human control. The autonomy of these systems is expected to be enhanced with intelligence gained from machine learning. As a result, intelligence-based software is expected to be increasingly used in airplanes and drones. It may also be utilized in airports and to manage air traffic in the future. Learning-enabled autonomous technology can also act as contingency management when used in safety applications, proactively addressing potential disruptions and unexpected aviation events. TRANSCEND was one of three projects chosen for the latest ULI awards. The others hail from Embry-Riddle Aeronautical University in Daytona Beach – researching continuously updating, self-diagnostic vehicle health management to enhance the safety and reliability of Advanced Air Mobility vehicles – and University of Colorado Boulder – investigating tools for understanding and leveraging the complex communications environment of collaborative, autonomous airspace systems. Florida Tech’s team includes nine faculty members from five universities: Penn State; North Carolina A&T State University; University of Florida; Stanford University; Santa Fe College. It also involves the companies Collins Aerospace in Cedar Rapids, Iowa and ResilienX of Syracuse, New York. Carroll and Bhattacharyya will also involve students throughout the project. Human operators are an essential component of aviation technology – they monitor independent software systems and associated data and intervene when those systems fail. They may include flight crew members, air traffic controllers, maintenance personnel or safety staff monitoring overall system safety. A challenge in implementing independent software is that engineers and operators have different interpretations of what makes a system “trustworthy,” Carroll and Bhattacharyya explained. Engineers who develop autonomous software measure trustworthiness by the system’s ability to perform as designed. Human operators, however, trust and rely on systems to perform as they expect – they want to feel comfortable relying on a system to make an aeronautical decision in flight, such as how to avoid a traffic conflict or a weather event. Sometimes, that reliance won’t align with design specifications. Equally important, operators also need to trust that the software will alert them when it needs a human to take over. This may happen if the algorithm driving the software encounters a scenario it wasn’t trained for. “We are looking at how we can integrate trust from different communities – from human factors, from formal methods, from autonomy, from AI…” Bhattacharyya said. “How do we convey assumptions for trust, from design time to operation, as the intelligent systems are being deployed, so that we can trust them and know when they’re going to fail, especially those that are learning-enabled, meaning they adapt based on machine learning algorithms?” With Bhattacharyya leading the engineering side and Carroll leading the human factors side, the research group will begin bridging the trust gap by integrating theories, principles, methods, measures, visualizations, explainability and practices from different domains – this will build the TRANSCEND framework. Then, they’ll test the framework using a diverse range of tools, flight simulators and intelligent decision-making to demonstrate trustworthiness in practice. This and other data will help them develop a safety case toolkit of guidelines for development processes, recommendations and suggested safety measures for engineers to reference when designing “trustworthy,” learning-enabled autonomous systems. Ultimately, Bhattacharyya and Carroll hope their toolkit will lay the groundwork for a future learning-enabled autonomous systems certification process. “The goal is to combine all our research capabilities and pull together a unified story that outputs unified products to the industry,” Carroll said. “We want products for the industry to utilize when implementing learning-enabled autonomy for more effective safety management systems.” The researchers also plan to use this toolkit to teach future engineers about the nuances of trust in the products they develop. Once developed, they will hold outreach events, such as lectures and camps, for STEM-minded students in the community. If you're interested in connecting with Meredith Carroll or Siddhartha Bhattacharyya - simply click on the expert's profile or contact Adam Lowenstein, Director of Media Communications at Florida Institute of Technology at adam@fit.edu to arrange an interview today.
Hurricane Preparedness Week reminds everyone to plan ahead
Hurricane season officially begins June 1, and now is the time to get ready for potential storms. Hurricane Preparedness Week, observed May 4 to 10, is a reminder for Louisiana residents to review their emergency plans, strengthen their homes, and prepare their families before a major storm arrives. “Preparedness saves lives, protects property and reduces recovery time,” said Carol Friedland, director of the LSU AgCenter LaHouse Research and Education Center. “By taking a few proactive steps now, families can be better protected during hurricane season.” Even though hurricane season does not begin for another month, starting your preparation now can save time and hassle later. The following are good projects you can start in the coming weeks to prepare for severe weather: Create and practice a family emergency plan and make sure it has updated contact information and evacuation information. Assemble an emergency supply kit or go through your existing kit to make sure nothing is expired or missing. Update it to make sure your entire family, including your pets, will have supplies during an evacuation. Review insurance coverage and understand flood risks. Some companies or policies have a waiting period before your coverage begins, so contact your agent or representative now. Strengthen your home against high winds and flooding. Do an overview of your home’s exterior and note any areas that look like they need repairs, like patches of missing shingles. Repairing these smaller problems now can reduce the likelihood of your home experiencing more severe damage during a storm. “Taking the time to check your home and yard early and getting bigger projects or repairs started ahead of hurricane season can save you a lot of stress later, especially when materials become harder to find right before a storm,” said Rubayet Bin Mostafiz, assistant director of research at LaHouse. Don’t wait until a storm is in the forecast. Start preparing today. Visit www.LSUAgCenter.com/LaHouse for more information, downloadable preparedness publications, and guidance tailored to Louisiana homes and families. The LaHouse Research and Education Center is dedicated to providing science-based solutions to improve home resilience, sustainability, and health. Through research, extension and education, LaHouse seeks to address the challenges of severe weather, with a particular focus on helping Louisiana residents protect their homes and communities from natural disasters. Article originally posted here.
'Chemtrails' and other climate conspiracies: Florida Tech's expert sets the record straight
When Florida Today columnist Tim Walters wanted to 'clear the air' about a popular conspiracy theory, he connected with Michael Splitt, an assistant professor at Florida Institute of Technology's College of Aeronautics with a focus on meteorology. The "chemtrail" conspiracy follows the erroneous belief that condensation trails (contrails) that trail behind jets are actually being used on a large scale to manage radiation and combat global warming. In the column, Splitt argued against the conspiracy by explaining what might happen if that level of "climate engineering" was actually going on. I recently wrote a column about the “chemtrail” conspiracy theory, and to say it caused quite a stir would be a serious understatement. My motivation for writing the piece came because there is a bill being looked at by the Florida legislature to address concerns of people who think the skies are being seeded by commercial airplanes with poisonous, weather-manipulating substances. Some of those raising concerns claim there are vague amorphous operatives in the federal government leading this charge. I decided I’d try to find answers, and I did so by asking someone credible in the field of weather sciences. Answers from climate expert Can the climate be altered by humans? The idea of trying to manipulate weather is called “climate engineering.” There is a form of this called solar geoengineering. “We've been doing things like this for decades in terms of, for example, fog management products. People have used this kind of methodology of adding things to the air to help get rid of fog, like the ice fog problem in Salt Lake City. So, there are places where people try to manage a local cloud layer,” Splitt said. However, it’s not done to a scale that would impact the country or globe. That’s where conspiracy theorists take climate engineering a step too far. There are those who say commercial airliners are spraying other substances like aluminum and barium (and other metallic) nano particles to reflect the sun's heat to reduce global warming. Splitt said if this were real, it might have the opposite effect. “When you have more contrails, it actually ends up warming the planet. The cirrus clouds created by aircraft and their reflective power isn't as much as let's say, the warming impact from below, from infrared radiation, so they end up being warmer." March 20 - Florida Today In the full column, Splitt also takes on other common misconceptions such as, "Why do some contrails last longer than others?" And, "Are ‘chemtrails’ steering, strengthening storms?" It's a worthwhile read for those interested in meteorology or conspiracy theories. Are you curious or looking to know more about those chasing clouds? Michael Splitt is available to speak with media. Contact Adam Lowenstein, Director of Media Communications at Florida Institute of Technology, at adam@fit.edu to arrange an interview today.
NASA Grant Funds Research Exploring Methods of Training Vision-Based Autonomous Systems
Conducting research at 5:30 a.m. may not be everybody’s first choice. But for Siddhartha Bhattacharyya and Ph.D. students Mohammed Abdul, Hafeez Khan and Parth Ganeriwala, it’s an essential part of the process for their latest endeavor. Bhattacharyya and his students are developing a more efficient framework for creating and evaluating image-based machine learning classification models for autonomous systems, such as those guiding cars and aircraft. That process involves creating new datasets with taxiway and runway images for vision-based autonomous aircraft. Just as humans need textbooks to fuel their learning, some machines are taught using thousands of photographs and images of the environment where their autonomous pupil will eventually operate. To help ensure their trained models can identify the correct course to take in a hyper-specific environment – with indicators such as centerline markings and side stripes on a runway at dawn – Bhattacharyya and his Ph.D. students chose a December morning to rise with the sun, board one of Florida Tech’s Piper Archer aircraft and photograph the views from above. Bhattacharyya, an associate professor of computer science and software engineering, is exploring the boundaries of operation of efficient and effective machine-learning approaches for vision-based classification in autonomous systems. In this case, these machine learning systems are trained on video or image data collected from environments including runways, taxiways or roadways. With this kind of model, it can take more than 100,000 images to help the algorithm learn and adapt to an environment. Today’s technology demands a pronounced human effort to manually label and classify each image. This can be an overwhelming process. To combat that, Bhattacharyya was awarded funding from NASA Langley Research Center to advance existing machine learning/computer vision-based systems, such as his lab’s “Advanced Line Identification and Notation Algorithm” (ALINA), by exploring automated labeling that would enable the model to learn and classify data itself – with humans intervening only as necessary. This measure would ease the overwhelming human demand, he said. ALINA is an annotation framework that Hafeez and Parth developed under Bhattacharyya’s guidance to detect and label data for algorithms, such as taxiway line markings for autonomous aircraft. Bhattacharyya will use NASA’s funding to explore transfer learning-based approaches, led by Parth, and few-shot learning (FSL) approaches, led by Hafeez. The researchers are collecting images via GoPro of runways and taxiways at airports in Melbourne and Grant-Valkaria with help from Florida Tech’s College of Aeronautics. Bhattacharyya’s students will take the data they collect from the airports and train their models to, in theory, drive an aircraft autonomously. They are working to collect diverse images of the runways – those of different angles and weather and lighting conditions – so that the model learns to identify patterns that determine the most accurate course regardless of environment or conditions. That includes the daybreak images captured on that December flight. “We went at sunrise, where there is glare on the camera. Now we need to see if it’s able to identify the lines at night because that’s when there are lights embedded on the taxiways,” Bhattacharyya said. “We want to collect diverse datasets and see what methods work, what methods fail and what else do we need to do to build that reliable software.” Transfer learning is a machine learning technique in which a model trained to do one task can generalize information and reuse it to complete another task. For example, a model trained to drive autonomous cars could transfer its intelligence to drive autonomous aircraft. This transfer helps explore generalization of knowledge. It also improves efficiency by eliminating the need for new models that complete different but related tasks. For example, a car trained to operate autonomously in California could retain generalized knowledge when learning how to drive in Florida, despite different landscapes. “This model already knows lines and lanes, and we are going to train it on certain other types of lines hoping it generalizes and keeps the previous knowledge,” Bhattacharyya explained. “That model could do both tasks, as humans do.” FSL is a technique that teaches a model to generalize information with just a few data samples instead of the massive datasets used in transfer learning. With this type of training, a model should be able to identify an environment based on just four or five images. “That would help us reduce the time and cost of data collection as well as time spent labeling the data that we typically go through for several thousands of datasets,” Bhattacharyya said. Learning when results may or may not be reliable is a key part of this research. Bhattacharyya said identifying degradation in the autonomous system’s performance will help guide the development of online monitors that can catch errors and alert human operators to take corrective action. Ultimately, he hopes that this research can help create a future where we utilize the benefits of machine learning without fear of it failing before notifying the operator, driver or user. “That’s the end goal,” Bhattacharyya said. “It motivates me to learn how the context relates to assumptions associated with these images, that helps in understanding when the autonomous system is not confident in its decision, thus sending an alert to the user. This could apply to a future generation of autonomous systems where we don’t need to fear the unknown – when the system could fail.” Siddhartha (Sid) Bhattacharyya’s primary area of research expertise/interest is in model based engineering, formal methods, machine learning engineering, and explainable AI applied to intelligent autonomous systems, cyber security, human factors, healthcare, explainable AI, and avionics. His research lab ASSIST (Assured Safety, Security, and Intent with Systematic Tactics) focuses on the research in the design of innovative formal methods to assure performance of intelligent systems, machine learning engineering to characterize intelligent systems for safety and model based engineering to analyze system behavior. Siddhartha Bhattacharyya is available to speak with media. Contact Adam Lowenstein, Director of Media Communications at Florida Institute of Technology at adam@fit.edu to arrange an interview today.
How authorship language helped catch a domestic terrorist – new podcast
In the latest episode of Writing Wrongs, hosts Professor Tim Grant and Dr Nicci MacLeod interview Dr Isobelle Clarke to unravel a case where forensic linguistics helped track down and convict a dangerous individual. Episode three, Imposters Tending to the Wild with Dr Isobelle Clarke, dives into the chilling case of Nikolaos Karvounakis, a self-proclaimed anarchist who planted a viable explosive device in Princes Street Gardens, Edinburgh, in 2018. Karvounakis, a Greek national, evaded capture for years, hiding behind online anonymity and extremist rhetoric. However, forensic linguists stepped in to analyse his anonymous blog posts, revealing patterns in his language that ultimately helped Police Scotland link him to the crime. The case not only demonstrates how linguistic evidence can be a powerful forensic tool but also raises crucial questions about the role of language analysis in modern terrorism investigations. On 11 January 2018, a suspicious cardboard box was discovered in a public seating area in Edinburgh’s Princes Street Gardens. After a controlled explosion, investigators determined the device could have caused serious harm had it detonated. With no immediate leads, the investigation stalled - until an anonymous blog post surfaced, claiming responsibility for the attack. The post, written in both English and Spanish, was linked to an eco-anarchist group called Individualists Tending to the Wild, a Mexican-based extremist organisation advocating violent action against technological progress. Crucially, the post included an image of the bomb’s interior, a detail only the perpetrator or law enforcement could have known. Police Scotland sought the expertise of Professor Tim Grant, who analysed the text, producing a linguistic profile that suggested the writer was neither a native English nor Spanish speaker - but rather someone influenced by another language entirely. Two years later, police identified Nikolaos Karvounakis as a suspect. Using comparative authorship analysis, Professor Tim Grant compared his online writings - including song lyrics from his rock band - to the manifesto. By dissecting word patterns, grammatical structures and stylistic quirks, he established that Karvounakis was the likely author. This evidence -alongside forensic meteorology, which linked photos of clouds in Karvounakis’ blog posts to the same weather conditions on the day of the crime - was used to secure a warrant and seize computers containing known writings by Karvounakis. To eliminate inevitable bias that would result from having worked the case for more than two years, Professor Grant invited Dr Isabelle Clarke onto the case as an independent forensic linguist. Using a version of the General Imposters Method, a technique similar to a police lineup but for language, Dr Clarke confirmed that the writing style in the blog post was the closest to Karvounakis’ known writings. Police Scotland put the evidence in the case, including the linguistic evidence, to Karvounakis, and secured a guilty plea. In February 2022, Nikolaos Karvounakis was sentenced to over eight years in prison under the UK’s Terrorism Act. Tim Grant, professor of forensic linguistics at Aston University, said: “The case highlights the growing importance of forensic linguistics in solving crimes, particularly in an age where digital anonymity combines with extremist ideologies. “It also highlights the how different types of language analysis can assist as a case moves through different stages of investigation.” Dr Nicci MacLeod, deputy director of the Aston Institute for Forensic Linguistics, said: “This episode offers listeners a behind-the-scenes look at the forensic methods that expose deception, identify threats and ultimately bring criminals to justice.” Dr Isobelle Clarke, a lecturer in security and protection science at Lancaster University and one of the first graduates from the campus-based MA Forensic Linguistics programme at Aston University, said: “It was great to be back at Aston University talking about the Karvounakis case for the Writing Wrongs podcast. “It’s an interesting case to highlight, as it shows how different types of language analysis can help with police investigations.” Writing Wrongs is available on Spotify, Apple Podcasts and all major streaming platforms. Listeners are encouraged to subscribe, share and engage with the hosts by submitting their forensic linguistics questions. Whether it’s about this case or broader forensic linguistic techniques, Professor Grant and Dr MacLeod welcome inquiries from listeners.
