Auburn Theses and Dissertations

Probabilistic Data Association of RADAR and LiDAR Measurements for Overtaking Maneuvers in Autonomous Racing

2026-09-15T20:58:03Z

Probabilistic Data Association of RADAR and LiDAR Measurements for Overtaking Maneuvers in Autonomous Racing Barksdale, Michaela This thesis compares the tracking performance of estimation done using cluttered RADAR velocity measurements to estimated velocity based on LiDAR position measurements alone. The lead vehicle state estimation was done with a Probabilistic Data Association (PDA) Filter. This is a modified Kalman Filter that involves calculating the probability each detection originated from the target being tracked and using a combined measurement innovation. This prevents the track divergence that often occurs when using a filter, such as the Nearest Neighbor filter, that selects one measurement to use based on mahalanobis distance. Once the lead vehicle was tracked via the PDA filter, the state estimate was used to inform an attempt at an overtake via model predictive control. With the rise in demand for vehicles with autonomous driving capabilities, the niche of autonomous racing research has also grown. One of the main tasks to perform in a racing scenario is an overtaking maneuver. To do this successfully, the opponent’s vehicle position and velocity need to be known. This tracking is done using active perception sensors on board the ego vehicle, including LiDAR and RADAR. RADAR is capable of measuring relative velocity, where LiDAR is not, but its raw detections are noisy and cluttered. Testing of the developed autonomy pipeline was done on both simulated and real Turtle- Bot4s. The PDA filter implementation was tested with a Lincoln MKZ and Kia Optima as well. This testing was meant to validate the method of vehicle tracking for future implementation on an autonomous indy racecar. LiDAR outperformed RADAR when estimating lead vehicle position in all experiments. The velocity tracking results suggest that there may be a benefit to having a direct measurement of range rate from the RADAR sensor. This difference was not significant in the case of the TurtleBot4. The result of the full-scale vehicle tracking did indicate a greater improvement in velocity tracking at slow speeds. Highway and racing speeds were not tested.

Advancements in Bacterial Diseases Affecting Channel Catfish (Ictalurus punctatus): Vaccine Development, Virulence Assessments, and Coinfection Interactions

2026-09-01T16:04:03Z

Advancements in Bacterial Diseases Affecting Channel Catfish (Ictalurus punctatus): Vaccine Development, Virulence Assessments, and Coinfection Interactions Harrison, Courtney Columnaris disease, caused by columnaris-causing bacteria (CCB), is a constant threat to farmed fish species. Several types of columnaris vaccines have shown various levels of protection in multiple fish species; however, it is hypothesized that due to CCB host associations, development of a vaccine that targets specific host species could elicit a strong immune response and provide protection. The first study of this dissertation investigated rifampicin-resistant CCB strains as potential live-attenuated vaccine candidates in channel catfish (Ictalurus punctatus) and Nile tilapia (Oreochromis niloticus). Five Flavobacterium covae (ALG-00-530, FBCC-CC-12K, ML-17-22A, IPRS-15-49, and ALG-15-231) and two Flavobacterium oreochromis (Costa Rica 04-02-TNT and R18-27) mutant strains demonstrated attenuation in channel catfish and Nile tilapia, respectively. Comparative genomic analysis of parent and mutant strains predicted deleterious rpoB mutations within 8/10 rifampicin-resistant mutant genomes. Other mutations affecting gene transcription, nutrient acquisition, and essential physiological pathways were also predicted as deleterious, suggesting attenuation is multifactorial. The second study assessed the efficacy of the five F. covae mutant isolates generated from the first study as live-attenuated vaccines in channel catfish. The vaccines were administered to fish via a static immersion for 30 minutes, then the fish were maintained for up to 8 weeks post-vaccination. Fish were then exposed to corresponding virulent parent strains and cumulative percent mortality and relative percent survival was determined. Two mutant strains (FBCC-CC-12K and ALG-15-231), exhibited variable protection; however, serum IgM antibody titers did not correlate with protection observed in channel catfish, and vaccinated fish antibody titers were comparable to sham-vaccinated fish. Protein profiles revealed conserved antigenic patterns between the parent and mutant strains. This study demonstrated protective efficacy of F. covae mutants is strain-specific and trial dependent, with F. covae FBCC-CC-12K and ALG-15-231 mutants warranting further investigation and optimization as live-attenuated vaccines. The third study involved virulence screening of yellow-pigmented bacteria recovered from diseased farmed catfish. Whole genome sequencing of isolates revealed the identities of one Chryseobacterium gleum, four C. mucoviscodosis, one C. cucumeris, one Empedobacter brevis, and one Flavobacterium saliperosum. The E. brevis isolate demonstrated high mortality and clinical signs of disease following intraperitoneal injection in channel catfish. Koch’s postulates were fulfilled in a replicated study, and the median lethal dose was estimated. A co-infection study with E. brevis and F. covae demonstrated significant mortality in the co-infection group compared to the single infection group of F. covae, but no significant difference in mortality between the co-infection group and single infection of E. brevis. This study demonstrated the exacerbated effect on mortality and disease following co-infection with E. brevis and F. covae. The final study assessed growth and proteomic responses of F. covae (Fc), virulent Aeromonas hydrophila (vAh), and Edwardsiella ictaluri (Ei) in co-culture. Growth metrics and secreted extracellular protein responses were assessed for three bacterial co-culture combinations. Growth metric analysis revealed cooperative growth in co-culture of Fc + Ei, while competitive growth was observed in co-cultures of vAh + Ei and Fc + vAh. Differential abundance of virulence-associated secreted proteins provided insight into species-specific alterations in co-culture. Collectively, these studies advance the understanding of host-specific live-attenuated vaccine development against columnaris disease, identify other yellow-pigmented bacterial pathogens affecting farmed catfish, and provide insight to mechanisms that influence bacterial virulence during co-infection.

An ex vivo biomechanical study of equine neonatal rib fracture repair by different fixation methods

2026-08-27T13:41:06Z

An ex vivo biomechanical study of equine neonatal rib fracture repair by different fixation methods Newell, Kayla Introduction: Rib fractures result in significant morbidity and mortality of equine neonates. Various methods to repair these fractures have been described but biomechanical testing in immature equine neonatal bone has not been performed. The study objective was to compare strength, stiffness, and mode of failure for four different methods of equine neonatal rib fracture repair. Methods: Equine neonatal cadavers (< 2 weeks of age) without history of dystocia or thoracic trauma present were used for this study. Ribs three through six on both left and right side were randomly allocated to one of four repair groups: nylon cable tie (NCT), modified reconstruction plate (MRP), string of pearls plate (SOP), or rib clip (RC). Intact ribs were first fractured under three-point bending to induce a simple, transverse fracture. The fracture was then reduced and repaired according to the assigned repair group. Repaired ribs were then tested to failure under single-cycle, four-point bending. Strength (maximum force, Fmax, and force at failure, Ffail) and mode of failure were recorded. Stiffness (N/mm2) was calculated. Data were analyzed using linear mixed models Results: In intact ribs, no significant differences in Fmax (P=0.46), Ffail (P=0.35), or stiffness (P=0.90) were observed between groups. Among repair groups, Fmax, Ffail, and stiffness differed significantly (P<0.0001). NCT repairs had significantly lower Fmax and Ffail than MRP and SOP, while MRP and SOP had greater stiffness compared to NCT and RC. NCT and RC groups failed at the original fracture site, whereas MRP and SOP failed mostly due to screw pullout. Conclusion: Rigid fixation with MRP or SOP provides stronger fracture repair than NCT or RC when equine neonatal ribs are tested under single-cycle load to failure. Further investigation of both rigid and non-rigid repair methods with cyclic loading is warranted.

Laboratory and Computational Evaluation of Mast Arm Foundations Under Torsion and Lateral Loading

2026-08-27T12:46:38Z

Laboratory and Computational Evaluation of Mast Arm Foundations Under Torsion and Lateral Loading Rojas Esparza, Olga Lucia Drilled shaft foundations are commonly used to support traffic signal mast arm structures because of their ability to resist large lateral, overturning, and torsional loads. Although current design procedures generally evaluate lateral and torsional loading independently, field observations and previous research indicate that these loading modes may interact, particularly under extreme wind events. This study investigates the behavior of drilled shaft foundations subjected to pure torsional, pure lateral, and combined lateral-torsional loading through a comprehensive program of full-scale testing and numerical modeling. Seven full-scale load tests were performed on two 36-in-diameter drilled shafts constructed within a controlled geotechnical chamber at Auburn University. The testing program included pure torsional loading, pure lateral loading, combined lateral-torsional loading, and sequential loading cases in two distinct phases. In Phase I, shafts were constructed in place through excavation and concrete placement. In Phase II, the shafts were exhumed, reinstalled, and surrounded by recompacted sand to evaluate the influence of construction method and shaft-soil interface conditions. Load, displacement, rotation, and strain measurements were collected throughout the testing program. The experimental results showed that torsional and lateral behavior are strongly influenced by loading sequence and shaft installation conditions. Pure torsional tests in both phases yielded similar ultimate torsional capacities of approximately 145-148 kip-ft; however, the recompacted Phase II configuration exhibited substantially higher initial torsional stiffness. Combined lateral-torsional loading reduced the mobilized torsional stiffness and altered the progression of resistance mobilization, requiring larger rotations to achieve the same torque levels observed under pure torsion. Sequential loading tests indicated that prior torsional loading reduced lateral stiffness and delayed the mobilization of resistance, although differences in ultimate lateral capacity were less pronounced. Numerical models were developed using FB-MultiPier and back-calculated against the measured load-displacement and torque-rotation responses. The best-fit estimation demonstrated that stiffness-related parameters, particularly shear modulus and horizontal subgrade reaction, were highly sensitive to construction effects, loading history, and stress path. Strength-related parameters, including unit weight, friction angle, and torsional shear resistance, exhibited comparatively smaller adjustments. The results demonstrate that conventional uncoupled design approaches do not adequately represent the behavior of drilled shafts under realistic loading conditions and highlight the need to consider construction effects, shaft-soil interface conditions, and load interactions in foundation design for mast arm structures.