Artificial Intelligence in Engineering

How UCF Is Impacting Various Engineering Disciplines with AI

From transforming wireless communication to designing more intuitive neural interfaces, UCF is at the global forefront of artificial intelligence innovation across multiple engineering disciplines. By using responsible AI enhancements, the university aims to continually discover new ways in which it can provide solutions and overcome challenges in the following disciplines:

• Computer Systems Design and Architecture
• Bioengineering and Computational Modeling
• Digital Twin, Virtual Reality, Augmented Reality and Interactive Visualization
• Brain-machine Interfaces and Robotics
• Signal Processing, Telecommunications and Wireless Networking

Computer Systems Design and Architecture

AI has significantly changed the design and architecture of modern computing systems, from personal devices to data centers. Self-learning algorithms can now automate and enhance processes by using techniques like reinforcement learning and neural architecture search. These methods create optimized chip layouts personalized to specific performance goals and manufacturing constraints. This allows much more rapid prototyping.

Additionally, AI facilitates self-optimizing computing architectures that continuously adapt to evolving user needs and workload demands. Known as adaptive computing, this approach uses techniques like predictive load balancing, just-in-time compilation and dynamic resource allocation to maximize efficiency.

Ronald DeMara, a Pegasus Professor in UCF’s Department of Electrical and Computer Engineering, specializes in optimizing computer system design and architecture through AI and machine learning. His research focuses on the intersection of circuits, computer architecture and new devices.

DeMara is exploring the use of a new type of device called a “probabilistic bit” for computing applications. The probabilistic bit is based on a spintronic device called a spin hall which affects magnetic tunnel junctions. It has a tunable threshold that allows the switching behavior to be tuned electrically without having to fetch instructions or data from memory — reducing energy consumption.

DeMara’s probabilistic bit device exhibits the potential to be used in analog computations to replace traditional Boolean circuits with floating point data paths. This could enable simpler, lower wiring count and more robust circuits that can handle intermittent power sources.

His goal is to minimize resource usage so these computing paradigms are suitable for small, inexpensive Internet of Things (IoT) devices. This pioneering research aims to push the limits of CMOS scaling and enable intelligent edge computing through energy-efficient analog computing approaches.

Bioengineering and Computational Modeling

Artificial intelligence is pivotal in analyzing massive, multidimensional biological datasets in bioengineering to reveal insights for advancing personalized medicine, medical devices and clinical diagnostics.

For example, AI techniques help construct detailed computational models that accurately simulate cardiovascular function and disease progression in digital twins. Researchers are able to utilize these models to customize treatments and predict outcomes based on a patient’s unique physiology.

Self-learning algorithms are now capable of solving complex problems that have long hindered traditional computational methods, presenting opportunities to enhance efficiency, sustainability, automation and innovation across industries.

Additionally, machine learning algorithms can process anatomical scans and biomarkers to rapidly diagnose conditions, identify optimal drug regimens or design tailored medical devices like stents and prosthetics. AI is also speeding up the screening of new pharmaceutical compounds.

UCF’s biomedical engineering program has been spearheading life-saving AI interventions for cardiovascular-related surgeries and congenital heart diseases. Under the leadership of UCF’s Pegasus Professor, Alain Kassab — trustee chair professor and lead coordinator of the biomedical engineering program — the team in his lab has successfully integrated AI-enhanced modeling to improve surgical planning, optimize device deployment and customize therapies. This has reduced adverse outcomes in patients with complex heart conditions.

Digital Twin, Virtual Reality, Augmented Reality and Interactive Visualization

Artificial intelligence has changed the landscape of simulated environments by introducing physics-based behaviors, predictive analytics and interactive responses that far surpass scripted systems. This transformation is enhancing extended reality (XR) applications across training, design and more.

For instance, AI gives digital twin simulations the ability to mirror real-world systems in action, forecast future outcomes and dynamically adapt to changes — providing invaluable analytics for complex processes like smart cities and supply chains. Furthermore, in virtual reality (VR)/augmented reality (AR), AI facilitates multi-sensory environments with near-natural immersive experiences.

UCF Agere Chair Professor, Carolina Cruz-Neira, is one of the researchers leading the way in the development of extended reality (XR) technology. Her Synthetic Reality Lab focuses on AI-enhanced simulations and interactive visualizations for experiential learning.

Cruz-Neira’s work includes collaboration with both government and industry partners to develop prototype solutions with tangible deliverables — VR welding simulators, field engineering training, interactive anatomy visualizations and rapid scenario generation for military training.

Currently, Cruz-Neira is focused on bringing social interaction elements into VR/AR and understanding how these technologies can be tailored to users’ specific roles within an application. For example, the way a medical visualization app is presented to a classroom of students would look different from how it’s presented to a practicing surgeon. By mirroring complex natural phenomena, AI is allowing for unprecedented applications of extended reality technology.

Brain-Machine Interfaces and Robotics

Artificial intelligence advancements in biotechnology, neurology and robotics have rapidly opened new doors of possibility. These advancements are driving remarkable progress in two key areas — neural interfaces that connect the human brain to computers and key robotics developments that can assist humans by operating with increased autonomy in the real world.

In brain-computer interfaces, AI analyzes neurological signals to interpret a user’s intended movements — facilitating naturalistic touch and proprioceptive feedback for enhanced usability. This allows more seamless control of assistive devices like motorized wheelchairs or bionic limbs using one’s mind. Similarly, in robotics, deep learning algorithms empower machines to perceive, adapt to and navigate within multi-faceted environments without explicitly programmed instructions.

Assistant Professor Mohsenz Rakhshan,of UCF’s Cognitive Neuroscience department, specializes in decoding brain activity and control assistive devices by utilizing AI. His Computational Neurophotonics Lab develops noninvasive, wearable systems that measure neurological signals to drive wheelchairs, computer cursors, robotic limbs and other rehabilitative technologies. Rakhshan’s team works closely with local clinicians to optimize neural interface technology for those with mobility impairments.

UCF’s research demonstrates how the combination of artificial intelligence with biotechnology and robotics can strengthen symbiotic relationships between humans and machines. For those who are neuro-atypical, disabled or injured, this progress is life altering.

Signal Processing, Telecommunications, and Wireless Networking

Artificial intelligence is driving major advancements in how data is processed, communicated, and transmitted across telecommunication networks. By enabling real-time analytics and automation, AI techniques like machine learning improve network efficiency, reliability, security and latency.

For example, AI provides adaptive signal processing that optimizes data transmission in response to live network conditions. This allows wireless systems to adjust frequencies, power levels and communication protocols to reduce interference and boost throughput.

One way in which AI enhancements have produced modernization in signal processing is through the automation of vehicles. At UCF, Professor Yaser Fallah specializes in enhancing wireless connectivity for autonomous and connected vehicles through artificial intelligence. His lab focuses on using cooperative AI, which involves learning-based perception and decision-making/planning for automated driving.

Specifically, Fallah’s team has contributed algorithms that have been adopted into vehicle communication standards used by major automotive manufacturers. They work on vehicle-to-vehicle and vehicle-to-infrastructure communication to offer advanced driver assistance systems. This looks like using reinforcement learning to train vehicles to demonstrate helpful behaviors such as slowing down to allow merging. The goal is safe and efficient cooperative autonomous driving.

As urbanization progresses, UCF will continue to develop AI enhancements that support future telecommunication networks.