Search for: All records

This project addresses the urgent need for inclusive and scalable robotics training in architecture, engineering, and construction (AEC) through the integration of artificial intelligence (AI) and extended reality (XR) technologies. In collaboration with three Minority Serving Institutions (Florida International University, Arizona State University, and University of Hawai‘i at Mānoa), we developed and tested immersive, adaptive learning environments that personalize robotics education for diverse student populations. These efforts include a VR-based curriculum for industrial robotics, an AR curriculum for environmental sensing technologies, and an overarching Robotics Academy framework that promotes open knowledge exchange and workforce connectivity. By combining real-time performance analytics, natural language processing, and biometric inputs, our systems support individualized learning paths and help mitigate algorithmic bias. This research advances equitable access to robotics education and provides a replicable model for technology-driven workforce development in the AEC sector. Ongoing evaluation demonstrates improved learner engagement, accessibility, and cross-platform skill transferability. 

Peterson, E., Bogosian, B., Vassigh, S. (2025). T. In Proc. 113th ACSA Annual Meeting (pp. ##-##). In press. 

Abstract We present a simulation of the time-domain catalog for the Nancy Grace Roman Space Telescope’s High-Latitude Time-Domain Core Community Survey. This simulation, called the Hourglass simulation, uses the most up-to-date spectral energy distribution models and rate measurements for 10 extragalactic time-domain sources. We simulate these models through the design reference Roman Space Telescope survey: four filters per tier, a five-day cadence, over 2 yr, a wide tier of 19 deg², and a deep tier of 4.2 deg², with ∼20% of those areas also covered with prism observations. We find that a science-independent Roman time-domain catalog, assuming a signal-to-noise ratio at a max of >5, would have approximately 21,000 Type Ia supernovae, 40,000 core-collapse supernovae, around 70 superluminous supernovae, ∼35 tidal disruption events, three kilonovae, and possibly pair-instability supernovae. In total, Hourglass has over 64,000 transient objects, 11,000,000 photometric observations, and 500,000 spectra. Additionally, Hourglass is a useful data set to train machine learning classification algorithms. We show that SCONE is able to photometrically classify Type Ia supernovae with high precision (∼95%) to az> 2. Finally, we present the first realistic simulations of non-Type Ia supernovae spectral time series data from Roman’s prism. 

Abstract A large fraction of Type Ia supernova (SN Ia) observations over the next decade will be in the near-infrared (NIR), at wavelengths beyond the reach of the current standard light-curve model for SN Ia cosmology, SALT3 (∼2800–8700 Å central filter wavelength). To harness this new SN Ia sample and reduce future light-curve standardization systematic uncertainties, we train SALT3 at NIR wavelengths (SALT3-NIR) up to 2μm with the open-source model-training softwareSALTshaker, which can easily accommodate future observations. Using simulated data, we show that the training process constrains the NIR model to ∼2%–3% across the phase range (−20 to 50 days). We find that Hubble residual (HR) scatter is smaller using the NIR alone or optical+NIR compared to optical alone, by up to ∼30% depending on filter choice (95% confidence). There is significant correlation between NIR light-curve stretch measurements and luminosity, with stretch and color corrections often improving HR scatter by up to ∼20%. For SN Ia observations expected from the Roman Space Telescope, SALT3-NIR increases the amount of usable data in the SALT framework by ∼20% at redshiftz≲ 0.4 and by ∼50% atz≲ 0.15. The SALT3-NIR model is part of the open-sourceSNCosmoandSNANASN Ia cosmology packages.