Gregory C. Lewin

Designing a spatially aware, automated quadcopter using an Android control system

Intelligence gathering is a critical component of military operations. Unmanned aerial vehicles (UAVs) have become an increasingly useful tool due to their surveillance and reconnaissance capabilities. However, the use of many of these vehicles is limited to outdoor environments because of their size and reliance on Global Positioning Satellites (GPS). Knowledge of indoor environments is important so that the risk of entering an unsafe or unknown building can be minimized. This paper describes the development of a spatially aware, autonomous quadcopter that uses an Android control system and functions indoors. The system consists of a laser rangefinder for sensory input, a IOIO microcontroller for data communication across platforms, an autopilot system (APM) for flight control, and an Android phone for mission control. The Android Control and Sensor System (ACSS) is currently being developed by the Department of Defense (DOD), MITRE, and academic partners, and will be integrated into the solution. To be considered autonomous, the quadcopter must be able to make a map from the data provided by the laser rangefinder, determine its own location and position in that map, and then execute a set of navigational commands from the Android control system. The success of this project is measured by the systems ability to travel autonomously while simultaneously creating a map and being aware of its location.

Designing a hexacopter for the collection of atmospheric flow data

Vertical profiles of temperature, pressure, relative humidity, wind speed, and wind direction in the atmosphere are typically collected using radiosondes attached to free-flying or tethered balloons. This method is inefficient when data are only required for the first hundred feet above the ground. Free-flying balloons and the attached payload drift away from the launching location and are often not recovered. Tethered balloons require large amounts of helium and become unstable with increased winds, and inflating balloons takes an extended period of time and requires a skilled team. The scope of this project is to eliminate the impracticalities of balloon-based measurement systems by creating a recoverable, versatile, user-friendly unmanned aerial vehicle (UAV). The project requires development of a flight-control system, a data-collection system, and a communications and user interface. The development of the flight-control system involved researching autonomous flight controllers, followed by the construction, prototyping, and tuning of a hexacopter. Creating the data collection system required researching environmental sensors and determining the effects of the copter motion on sensor performance. The designed communications interface incorporated realtime data flow and local storage on the copter. The final product will be an autonomously flying hexacopter which can collect accurate weather-related data within the lowest 1000 feet of the atmosphere.

Development of an autonomous multi-rotor copter for collecting atmospheric data near the ground

Vertical profiling of atmospheric data for the first several hundred meters above ground level is a powerful mechanism by which environmental scientists conduct research and development. Accurate measurements of atmospheric variables such as wind, temperature, and relative humidity, are essential to the many applications of this type of environmental research. The current methods of taking measurements by free-flying or tethered balloons can be inconsistent and unreliable. Beyond the scientific implications, balloons are also unpractical for large research projects that demand multiple profiles at high spatial and temporal resolution. Free-flying balloons and the attached sensors are often not recovered after use, and they can be costly to replace. Tethered balloons can be reused, but the cost of helium as well as the time and work required to operate make them undesirable for repetitive observation. The purpose of this project is to develop an unmanned aerial vehicle (UAV) that can collect atmospheric data for the purpose of generating vertical profiles of the first several hundred meters above the ground. This overall goal for the project includes not only the development of the vehicle itself, but also a data collection system and a mechanism for users to interface with the data both during and after data collection. The system is designed to be pre-programmable and autonomous, providing researchers the precision and control required to generate accurate profiles of the lower atmosphere. It will also be reusable, offering a clear advantage over the traditional balloons, and customizable to fit varying research needs. The system in its current form is a hexacopter built on the open source ArduCopter platform. It is outfitted with an extensive sensor package to measure wind speed and direction, temperature, humidity, and pressure. Data are logged directly to an onboard SD card, and additionally transmitted wirelessly to a laptop ground station when the vehicle is in use. When this team took control of the project, much of the hardware was already in place. As such, the teams efforts this past scholastic year were 1) the development of a robust information system that will allow users to both see the data in real-time in a useful format and easily manage and process the data after they are measured, and 2) further stabilization of the copters flight while hovering at a fixed altitude. In its final form, the copter and its counterpart information system will provide environmental scientists with a powerful tool that will both simplify and augment the process of low-altitude atmospheric research.

Designing a spatially aware, autonomous quadcopter using the android control sensor system

Gathering information for intelligence, surveillance, and reconnaissance (ISR) poses a risk to the human operators, namely the United States military and intelligence sectors. An autonomous drone that can perform advance ISR of enclosed spaces will significantly impact a variety of safety critical applications, including search and rescue. Current systems are limited to outdoor environments with access to global positioning systems (GPS) and are typically expensive, with custom engineering and proprietary interfaces. Our aim is to create indoor capability and utilize commercial off-the-shelf (COTS) subsystems to reduce cost and improve flexibility for diverse applications. The goal of this project is to develop a proof of concept design for a quadcopter that will create a map of an unknown, indoor space. We must develop a simultaneous localization and mapping (SLAM) algorithm for the quadcopter to create the map autonomously. The problem of both building a map of an unknown space and localizing within that space is termed SLAM. SLAM is frequently referred to as a chicken-and-egg problem, since accurate mapping requires knowledge of location, and vice versa. A SLAM algorithm must probabilistically relate environmental sensors and utilize a probabilistic motion model to converge to a most likely map of the environment and position of the robot. This project has four major parts: hardware, which includes integration of the sensors, quadcopter, and Android phone; command and control; the SLAM algorithm, which will run without a GPS; and a mobile application for viewing usable maps. We found both localization and mapping algorithms are adept at operating separately within a GPS obscured environment. Future steps include combining the localization and mapping algorithms into an optimized SLAM algorithm that will run efficiently on the Android phone.

Automated lighting system for park pathways

Smart park lighting has the potential to save energy and improve safety within the City of Charlottesville. Standard park lights remain lit throughout the night regardless of whether or not a user is present, which causes unnecessary energy waste. The City of Charlottesville has expressed an interest in retrofitting some of their park lights with an automated lighting control system. The automated system uses motion sensors to detect when users are on the path. The system turns the lights on when users are present, and turns the lights off otherwise. The goal of this project is to develop a long-term automated control system prototype for the City of Charlottesvilles path lights and to test it in a City of Charlottesville park. First and foremost, the system needs to maintain the safety of path travelers while saving energy. Additionally, the system and installation need to be relatively low cost. Trenching between poles to lay the underground wiring is the biggest installation expense for standard light poles. To avoid additional trenching costs, the poles will communicate wirelessly with their neighbors to turn on the lights of surrounding poles. The automated lighting system uses passive infrared (PIR) sensors located on each light pole to detect users on the path. A microcontroller handles the control logic of reading the sensor and determining when a light is turned on or off and when to wirelessly tell neighboring poles to turn on. The lights remain on for a specified timeout period unless motion is detected again. Parameters such as the timeout period and the number of neighboring poles that turn on after motion is detected are wirelessly configurable via a laptop connected to a wireless transceiver. These parameters will be thoroughly tested to ensure sufficient lighting is provided to meet user safety needs. The final deliverable is a reproducible automated lighting prototype with wirelessly configurable parameters installed in a Charlottesville park.

Wind data collection techniques on a multi-rotor platform

Meteorologists require reliable methods of obtaining high-quality atmospheric data, such as temperature, humidity, pressure, and wind velocity, to better understand and predict weather phenomena. Tethered weather balloons and ground towers are the current standards for research-grade atmospheric profiling. However, weather balloons and ground towers suffer from several disadvantages including high cost, low mobility, and labor intensive setups. Due to their dynamic nature, unmanned aerial vehicles (UAVs) do not suffer from those disadvantages.

AirChat: Ad hoc network monitoring with drones

FireChat is a mobile messaging application that uses a mesh network topology to communicate between users through a cellular, WiFi, or Bluetooth Low Energy connection. FireChat has seen global growth for use in crowded events such as protests. The team has worked in conjunction with the MITRE Corporation to develop a system that attaches to a quadrotor and monitors nearby connections to gather metadata on the FireChat. This information is analyzed to provide insight into the structure of large groups or specific devices using FireChat. The WiFi Pineapple Nano, Ubertooth One, and HackRF One were tested for monitoring FireChat. From these three, the Pineapple was selected to monitor WiFi and the Ubertooth was chosen to monitor Bluetooth. The devices were evaluated based upon their ability to provide useful data, meet size, weight, and power requirements, and function across all platforms on which FireChat can be installed. While the Ubertooth provided some relevant data about the use of FireChat, it was inconsistent so Bluetooth development was discontinued. The Pineapple, however, successfully collects the MAC address and connected routers of the FireChat devices. A Raspberry Pi runs scripts to control the Pineapple to scan WiFi networks and collect FireChat data. Further, the time and location will be saved based on timing of received data and location of the quadrotor. This data is sent to a server using the Particle Electron, a 3G cellular communication board, where further analytics are performed. The analysis provided to the sponsor is able to identify the location of certain FireChat users, their progression over time, and common communication networks in which they participate.

Using image data for Quality Assurance in Additive Manufacturing

Additive Manufacturing (AM) processes currently suffer from high failure rates, which can amount to immense danger when the AM parts are incorporated in complex systems used by the public, i.e. cars, airplanes, or construction. Most current AM Quality Assurance (QA) methods are conducted post production, though ultrasonic testing may also be conducted while a part is being built. The focus of the project is to build a non-destructive post-production verification process that uses image data to render 3D models of the actual build, which will aid users in determining the quality of a part. The visualization process presented here will automate some labor intensive tasks and ultimately help users easily determine how print parameters affect the quality of a build by 1) collating and cleaning image data from actual builds, 2) storing data efficiently, and 3) creating visualizations of the images. A crucial outcome of this research is the ability to determine the quality of a final printed object without destroying it for QA inspection.

Designing an autonomous soil monitoring robot

Through the monitoring of soil conditions land managers can respond rapidly to mitigate adverse events, such as extreme weather or ongoing drought. However, without an extensive system of sensors, gathering information over a large field takes an exorbitant amount of time. This mass collection of soil data would allow farm managers to study time-lapsed trends and variables within a particular region to provide quick assessment of land conditions. Currently, the client uses a bulky handheld wireless soil sensor to measure moisture content and temperature. To take measurements, the client must walk to the coordinates of interest, clear the ground of vegetation, manually insert the probe into the ground, and log the reading. The team is designing an autonomous soil monitoring rover to expedite data collection and reduce labor. The rover will be able to autonomously navigate through a field several acres in size and avoid obstacles. It will gather data on soil moisture and temperature at a set of given waypoints and relay the information back to the farm manager. Constructed with a custom welded steel frame, the first rover prototype will be a four-wheeled vehicle with front wheel drive. The vehicle will be equipped with a Stevens Hydra Probe II mounted to a linear actuator. Navigation will be handled using a GPS and wheel encoders. When completed, the rover will allow the land manager to analyze trends between soil data and pasture health, providing an accurate snapshot of a field.

A low-cost chemical sensor for fixed security applications

A low-cost infrared sensor that uses room temperature pyroelectric detectors integrated with bandpass filters to provide low-resolution spectral scans of the absorption characteristics of hazardous chemicals was developed for fixed security applications. The sensor provides fast (1 s) and continuous monitoring, detection, and identification capabilities. A unique detection and identification algorithm that uses non-linear computation techniques to account for the exponential nature of optical absorption was developed. Chemical detection and identification is achieved by matching the recorded sensor response vector to an updatable signature library that currently includes the signatures of 14 chemicals. The sensor and algorithm were tested by introducing methanol vapor at optical depths between 225 - 270 ppm-m. Using 1 s signal samples obtained during approximately 20 min. test, resulted in no false positive alarms and 3.4% of false negatives. All false negatives were shown to be due to misidentification of methanol as isopropanol, which is spectrally similar to methanol. By grouping isopropanol with methanol the rate of false negatives was reduced to 0%. Results of the same test using a 30 s signal integration time resulted in no false positive and no false negative alarms.