The purpose of this experiment was to visualize and record the different rates of expansion for multiple gases as they are launched into the higher parts of Earth’s atmosphere with a High-Altitude Balloon (HAB). The ideal gas law models the behavior of a gas that of which its molecules occupy no volume and have no intermolecular forces (IMF). It is a simple equation; however, it cannot model gases accurately. On the other hand, Van der Waals equation for non-ideal gases better resembles the behavior of a real gas as it includes what the ideal gas law lacks. To test this, we filled three syringes with three different gases to the same volume. We chose to test argon, helium, and nitrogen. We secured the syringes to a container, which served as the payload for the HAB. We also placed an altimeter, thermometer, and a barometric pressure sensor inside the container. Then, we connected the sensors to an Arduino to record each piece of data synced to a stopwatch that is displayed in the container on a screen. Finally, we secured a camera to the container facing the stopwatch and syringes to record the gasses’ volume. Because helium has the weakest IMFs out of the three gases, we believed helium would expand at a higher rate as atmospheric pressure decreases compared to the other gases. The results from our experiment serve as a good example of how far the behavior of real gases deviate from ideal gases modeled by the ideal gas law. Depending on how close our measured values reach the calculated values from the ideal gas law, we can predict which situations the ideal gas law can model the behavior of a particular gas relatively accurately.

Multitasking, such as walking and talking, is common for humans and other animals, yet animals are limited in how many behaviors they can perform simultaneously. The neural circuit mechanisms that limit multitasking are not well understood. Uncovering these mechanisms will help us understand how brains seamlessly combine behaviors. In turn, this will improve our understanding of some disorders, such as Parkinson’s Disease, wherein certain symptoms, like difficulties with walking and talking, are exacerbated during multitasking. I am investigating how neural circuits for known behaviors interact to limit multitasking in a genetic model system, the fruit fly Drosophila melanogaster. During the first year of my project, I developed a rig to record complex behaviors and I added optogenetic activation to the rig in order to activate neurons that drive one motor program during single and multitasking contexts. I am now preparing to conduct experiments with this rig, comparing the probability that a fly extends a wing (driven via optogenetic stimulation) while resting (single-tasking) vs. walking (multitasking). I expect that a fly will have a higher wing extension probability while single-tasking than while multitasking. The results of my project will help uncover how neural circuit interactions shape an animal’s ability to multitask.

In an ideal world, the electrical grid could fully decarbonize with just solar and wind as forms of energy generation. However, more traditional and firm forms of generation, such as natural gas or nuclear power, support the stability of the electric grid and lower the cost of transition to a net-zero carbon grid. To properly integrate emerging technologies, such as carbon capture and sequestration (CCS), as sources of firm generation, there needs to be an understanding of their economic behavior, and larger impact on other forms of electrical generation. CCS is of interest because the natural gas industry supplies the cheapest electricity and plays a major role in planning the energy transition. This project seeks to understand the economic viability of CCS to inform policy encouraging the deployment of emerging electricity resources. How do the investment costs associated with CCS need to change to result in significant buildout of natural gas plants with CCS? How does the increase of CCS buildout impact other forms of generation within a given system, and contribute to the overarching goal of creating a sustainable energy future? Using the MIT Energy Initiative's capacity-expansion model GenX, the investment costs associated with adding new natural gas plants with CCS and with retrofitting existing plants with CCS are varied in a sweep and the impact on the amount of added capacity of CCS and other forms of generation is analyzed. It has been found that an 80% reduction of the investment costs associated with CCS begins to promote the buildout of new CCS plants. Further investigation on the impact to other power sources, particularly battery storage, will provide insight into the impact of expanding CCS capacity on the rest of the system, with the anticipated result that increasing CCS buildout discourages the buildout of already established power sources.
 

In quantum computing, computer engineers require a method to control the logical state of a quantum bit (qubit). Unlike its classical counterpart, a qubit’s logical state is not defined by a binary and discrete voltage. The QT3 lab is developing a quantum testbed with a defect in diamond known as the nitrogen-vacancy (NV) center. We present an on-diamond antenna that is optimized to manipulate the electron spin state of an NV center, which defines the qubit. Applying a radiofrequency magnetic field equal to the energy difference between the two spin states of our qubit, also known as resonant excitation, enables control of this state. The strength of this field directly correlates to the frequency at which this manipulation may occur. That frequency is known as the Rabi frequency. This is important as we want this frequency to be faster than the state can undergo decoherence, where state information is lost to the environment of the qubit. We have designed and simulated an antenna using finite element analysis software, which will supply our field and be fabricated on the diamond surface. For its geometry we realized a coplanar waveguide with a shorted end shaped around the NV center, which optimizes the field strength at the NV center, power reflections, and area consumption. Preliminary fabricated samples have been mounted, wirebonded, and characterized using a vector network analyzer, and have shown behavior that aligns with simulated results. We expect to have the antenna fabricated on our single NV center testbed sample and achieve a Rabi frequency on the order of 10’s of MHz. Once this sample is fully integrated into our cryogenic system, it will enable us to expand control to multiple nuclear spin qubits from a single NV center, as a quantum register. The testbed will be accessible to researchers and educators.

As cloud computing and artificial intelligence continue to evolve, data centers increasingly rely on a variety of accelerators to enhance performance. Field Programmable Gate Arrays (FPGAs), known for their exceptional speed and flexibility, are emerging as a viable alternative to conventional processors such as CPUs, GPUs, and ASICs. Despite their potential, a key challenge in cloud-based FPGA deployment is efficient multi-tenancy, particularly due to the nascent stage of this technology in cloud environments. This study addresses this challenge by focusing on the emerging Disaggregated FPGA Architecture, which conceptualizes FPGAs as independent compute nodes. This approach facilitates smoother integration into data centers and offers a solution to prevalent integration challenges. The importance of this study lies in its potential to revolutionize the deployment and scalability of FPGAs in cloud environments, thereby significantly enhancing data center performance and efficiency. This not only contributes to the field of cloud computing by advancing FPGA integration strategies but also sets a new benchmark in resource optimization and scalability for cloud infrastructures. To underscore the effectiveness of Disaggregated FPGA architectures, the research first advocates for their strategic advantages in cloud settings. Following this, a detailed test plan is presented, comparing Disaggregated FPGA architectures with existing cloud FPGA frameworks. This comparison aims to validate the feasibility and scalability of the proposed architecture. Through this research, we contribute a novel perspective to FPGA deployment in cloud computing, offering a potential paradigm shift in data center technology and paving the way for more efficient, scalable, and adaptable cloud infrastructures.

Global Positioning Systems (GPS) technology plays a pivotal role in ensuring the safe and efficient navigation of drones by providing near real-time tracking of location and speed. The precision and reliability of GPS receivers are crucial for effective planning, sensing, and control applications in various domains. As Unmanned Aerial Vehicles (UAVs) continue to rise in demand and predominantly rely on GPS, minimizing the uncertainty in GPS performance becomes imperative. UAVs utilize the cost-effective nature of Micro-electromechanical Systems (MEMS) GPS receivers. This study identifies and analyzes GPS errors, specifically within consumer-grade MEMS receivers. The MEMS receivers are preferred for their low cost, low power, and low weight, making them ideal for integration into UAVs. Our methods include a series of controlled experiments in urban and semi-urban environments, encompassing varying weather conditions such as sunny and cloudy days. Static experiments evaluate GPS signal accuracy under stationary conditions, while dynamic experiments monitor GPS performance during drone flights. Our preliminary findings have shown a range of inaccuracies in GPS signal measurements. Horizontal signal accuracy varied from +/-1 to +/-14 meters, while vertical signal accuracy ranged from +/-3 to +/-12 meters. These results underscore the significance of further investigation to enhance GPS reliability, particularly in scenarios critical for UAV operations. In ongoing research, we are conducting more testing in other geographical locations and weather conditions to ensure the robustness of our conclusion. Additionally, we are developing environment-specific error detection algorithms utilizing the sensor fusion approach. Merging data from multiple sensors can reduce the uncertainty of an object's location, helping us when the GPS technology is not fully reliable. Our research contributes to advancing GPS technology capabilities, particularly for UAVs where accurate localization is important.