<div>Wherever we go, we need resources. Finding those resources in unmapped areas is an ever-present challenge. Nature provides many examples of systems that manage to nd the resources they need for growth, despite having little to no information about their environment. Emulating the resource-finding strategies of animals and insects has been, and continues to be, attempted in robotic systems, to varying degrees of success. However, borrowing strategies from plants is much less explored. This dissertation explores an attempt at distilling the resource-hunting methods of plant roots into a collective robotic system.</div><div><br></div><div><div>Utilizing low-power computing and wireless communication, the robots attempt to locate "resources" (radio beacons, in this case) in an unmapped area. They work collectively via extending and branching from each other. The results of this experiment show limited success, with the limitations primarily stemming from the wireless communication. Nonetheless, it is shown that a collective robotic system, emulating plant roots, can feasibly locate resources that display a gradient, with no map of the environment.</div></div>

There is a significant amount of ongoing research on developing multi-agent algorithms for mobile robots. Moving those algorithms beyond simulation and into the real world requires multi-robot testbeds. However, there is currently no easily accessible source of information for guiding the creation of such a testbed. In this thesis, we describe the process of creating a testbed at Purdue University involving a set of unmanned aerial vehicles (UAVs). We discuss the components of the testbed, including the software that is used to interface with the UAVs. We also describe the challenges that we faced during the setup process, and evaluate the UAV platforms that we are using. Finally, we demonstrate the implementation of a multi-agent task allocation algorithm on our testbed.

Unmanned aerial vehicles (UAVs), when paired with an onboard camera, have proven to be useful tools in many applications, including aerial photography, precision agriculture, and search and rescue operations. Likewise, UAVs capable of physically interacting with the environment have shown great potential to help people perform dangerous, or time-consuming tasks more safely and efficiently than they could on their own. However, due to onboard computation and battery life limitations and complex flight dynamics, using UAVs to physically interact with the environment is still a developing area of research. Considering these limitations, the primary goals of this work are to (1) develop a new UAV platform for aerial manipulation, (2) develop modular hardware and software for the platform to enable specific tasks to be performed autonomously, and (3) develop a visual target tracking method to enable robust performance of autonomous aerial manipulation tasks in unstructured, real-world environments. To that end, the design of the Interacting-BoomCopter UAV (I-BC) is presented here as a new platform for aerial manipulation. With a simple tricopter frame, a single additional actuator for generating horizontal forces, and lightweight, modular end-effectors, the I-BC aims to balance efficiency and functionality in performing aerial manipulation tasks, and is able to perform various tasks such as mounting sensors in hard-to-reach places, and opening small doors or panels. An onboard camera, force and distance sensors, and a powerful single board computer (SBC) enable the I-BC to operate autonomously in unstructured environments, with potential applications in areas such as large-scale infrastructure inspection, industrial inspection and maintenance, and nuclear decontamination efforts.

<div>Path planning is the generation of paths for the robots to navigate based on some constraints. Coverage path planning is where the robots needs to cover an entire work space for various applications like sensing, inspection and so on. Though there are numerous works on 2D coverage and also coverage using a single robot, the works on 3D coverage and multi-agents are very limited. This thesis makes several contributions to multi-agent path planning for 3D structures.</div><div><br></div><div>Motivated by the inspection of 3D structures, especially airplanes, we present a 3D coverage path planning algorithm for a multi-UAV system. We propose a unified method, where the viewpoints selection and path generation are done simultaneously for multiple UAVs. The approach is scalable in terms of number of UAVs and is also robust to models with variations in geometry. The proposed method also distributes the task uniformly amongst the multiple UAVs involved and hence making the best use of the robotics team. The uniform task distribution is an integral part of the path planner. Various performance measures of the paths generated in terms of coverage, path length and time also has been presented. </div>

<p>High precision magnetic levitation control methodologies during the manufacture of Organic light-emitting diode (OLED) displays are designed, manipulated, and experimentally validated in this thesis. OLED displays have many advantages over conventional display technologies including thinner, lighter, lower power consumption, higher resolutions, and greater brightness. However, OLED displays require tighter environmental conditions of the manufacturing processes without the introduction of vibration and contamination. For this reason, magnetic levitation is used to transport the displays attached on the carrier during the manufacturing process. This thesis addresses several critical problems related to implement the levitation control performance of the carrier's motion during the manufacturing process. </p> <p>Attractive magnetic levitation requires measurement of the airgap between the carrier and the levitation electromagnets. An algorithm for modeling the gap sensor installation errors was developed and subsequently used for controller development. A levitation controller only was initiated as the stationary point for optimal state feedback controller-observer compensator developed in this study. This optimal state feedback controller-observer compensator allows the carrier to be passed from support fixtures without the introduction of vibration. This controller was designed, and its levitation control performance confirmed with both simulation and experimental validation. To implement the levitation control performance of the carrier's motion, a second order notch filter and a first order low pass filter are designed to minimize the mechanical resonance and noise from the gap sensor, respectively. To reduce the sudden change of the levitation forces owing to the discrete allocation of the levitation electromagnets, a section control algorithm is developed; the sum of the levitation forces is equal to the weight of the carrier and the sum of the moment along the propulsion axis is equal to zero. </p> <p>Using the developed control strategies, the peak to peak variation of the carrier’s motion at a standstill was 50 µm. This same motion at low-speed 30 mm/s was 250 µm. While at high speed 300 mm/s was 430 µm. The relative improvement in the levitation control performance of optimal state feedback controller-observer compensator over the levitation controller only was a peak to peak attenuation of 50 µm at low-speed and 270 µm at high-speed. Most significantly while using optimal state feedback controller-observer compensator could be passed from support fixture to support fixture, i.e., through the deadzone, without mechanical contact or other manufacturing processes, inhibiting vibration. </p> <p>Having comparative simulation and experimental validation, the proposed control strategies were validated to improve the levitation control performance of the carrier under uncertain disturbance and sensor installation error, and it is expected to manufacture OLED displays with high productivity and low defect rate.</p>

<div>Sediment has a significant impact on social, economic, and environmental systems. With the need for an effective sediment management and monitoring system growing more important,</div><div>a method for precisely and reproducibly obtaining sediment samples that represent the actual environment is essential for water resource management and researchers across aquatic domains (such as lakes, rivers, reservoirs, mine drainage ponds, and wastewater lagoons).</div><div>Sediment sampling is usually carried out less frequently than water sampling because of the cost and labor involved. However, more frequent sediment sampling and an increase in the</div><div>range of the sampling area are necessary to more effectively monitor the ecosystem and water quality.</div><div>To fill this gap, robotic approaches for sediment sampling have been introduced. However, they are not tailored to a sediment sampling method and do not focus on the quality of</div><div>the sediment sample. Moreover, there are many challenges involved in developing such a sediment sampling system for the surface water of rivers, streams, lakes, ponds reservoirs, and lagoons. Thus, this study can be conducted to investigate to design and develop an uncrewed sediment sampling system for surface-water environments based on marine robot platforms that are capable of collecting intact sediment samples from a range of sediment types. As part of this study, an unmanned surface vehicle (USV) was used to deploy the underwater sediment sampler (USS) at the sampling locations. The USS adopted a core sampling method to collect the sediment samples. The specific requirements were integrated, taking into consideration the challenges posed by surface water and underwater environments, to design and develop an unmanned sediment sampling system.</div><div>The USV has two missions - deploying and positioning. Users can deploy the USV with the USS to the desired sampling area. Once the USV arrives, it has to maintain its position while launching the USS and during the sampling process. The USS also has two missions — launching and sampling. The USS must be a negative-buoyancy platform so it can reach the bottom and maintain its stability during sampling. To sample the sediment, the USS has to generate a sampling pattern. We defined and formulated challenges based on the missions of each platform.</div><div><div>The USV consists of three sub-systems; propulsion, launching, and monitoring system to accomplish missions. The propulsion system and launching system are necessary to accomplish deploying and positioning missions. The propulsion system is consists of two thrusters to navigate the USV. The launching system is to launch anchors for positioning and the USS for sampling. The monitoring system is to monitor and control other systems on-board via online video. The USS can generate sampling patterns based on three motions; linear, rotational, and hammering motion. We integrated servos, sensors, and mechanical components to generate three motions. The main system of the USS is completely waterproof, even for linear and rotational motion with enclosures, O-rings, and rubber bellows. Since the USS operates underwater, the water pressure causes the pressure difference between inside and outside the enclosure. We designed a pressure-equalizing system to compensate for the volume change because of sampling motions and pressure differences. Extensive field experiments were conducted to evaluate the proposed system. Users can monitor and control the system from the base station based on all data and images from each platform. The evaluation of the system is based on the data from sensors installed on each platform. Deploying and positioning missions of the USV can be shown based on the trajectory data. Launching and sampling missions of the USS can be validated based on depth, orientation, and reaction force data.</div><div>Contributions of the proposed unmanned sediment sampling system are, 1) It is the first unmanned system with a novel design to collect the less disturbed sediment samples even</div><div>from the inaccessible area and remove the potential risks of human-based sampling tasks, 2) We proposed and integrated a new sediment sampling pattern based on the sediment</div><div>sampling pattern analysis to increase the quality of sediment samples by minimizing disturbances, and 3) The proposed unmanned sediment sampling system is the first step toward the autonomous environmental monitoring system for more effective environmental monitoring.</div><div>This proposed system has many potential elements that can be a total solution for robotic environmental monitoring in addition to other features such as water sampling system, and various types of sensing system.</div></div>

<div>Cyber-Physical Systems (CPSs) consist of physical and computational components usually interconnected through the internet. This type of systems have found applications in robotic surgery, smart medical services, driverless cars, smart power grids as well as in modern homes and offices. For a CPS to function properly, a reliable and secure communications between the system physical and cyber elements is of utmost importance. Malicious attacks during control signals and output measurements transmission between the physical plant and the control center must be addressed, which is the main research problem studied in this thesis.</div><div><br></div><div>A novel robust observer was proposed to synthesize a combined controller-observer compensator for a class of CPSs with sparse malicious attacks and arbitrary disturbances. The compensator consists of a controller, a norm approximator, and an unknown input observer (UIO). The proposed observer was compared with a norm-based observer given in the literature to show its advantage. To further enhance the proposed observer's performance against arbitrary disturbances, design methods were given that use fictitious output measurements and error correcting code (ECC) approach. The design of the UIO was extended to a bank of UIOs in order to improve the observer's performance against sparse malicious attacks.</div><div><br></div><div>The proposed observer can be used in the design of UIO-based fault detection and isolation (FDI) algorithms as well as in the distributed fault-tolerant control of large-scale interconnected systems. The results of this thesis can be applied to the design of controller-observer compensators for CPSs with modeling uncertainties.</div>

<p>Due to their rapid technological advancement, mobile sensors such as unmanned aerial vehicles (UAVs) are seeing growing application in the area of multi-target tracking and identity management (MTIM). For efficient and sustainable performance of a MTIM system with mobile sensors, proper algorithms are needed to both effectively estimate the states/identities of targets from sensing data and optimally guide the mobile sensors based on the target estimates. One major challenge in MTIM is that a target may be temporarily lost due to line-of-sight breaks or corrupted sensing data in cluttered environments. It is desired that these targets are kept tracking and identification, especially when they reappear after the temporary loss of detection. Another challenging task in MTIM is to correctly track and identify targets during track coalescence, where multiple targets get close to each other and could be hardly distinguishable. In addition, while the number of targets in the sensors’ surveillance region is usually unknown and time-varying in practice, many existing MTIM algorithms assume their number of targets to be known and constant, thus those algorithms could not be directly applied to real scenarios.</p> <p>In this research, a set of solutions is developed to address three particular issues in MTIM that involves the above challenges: 1) using a single mobile sensor with a limited sensing range to track multiple targets, where the targets may occasionally lose detection; 2) using a network of mobile sensors to actively seek and identify targets to improve the accuracy of multi-target identity management; and 3) tracking and managing the identities of an unknown and time-varying number of targets in clutter.</p>

This dissertation describes the evolution of the different local magnetic field generating systems for independent actuation of multiple microrobots. A description of the developed hardware, system characterization tests, and experimental results are presented. The system is designed for automated control of multiple microrobots. Finally, sample micromanipulation tasks are demonstrated using the new microrobot design, showcasing its improved manipulation capabilities.<br>First, a mm-scale local magnetic field generating system designed for single layer coils is used to control 3.175 mm size N52 magnets as robots independently in the workspace. The controller used a set of local equilibrium points that were generated from a sequence of coil currents around the robots from one state to the next. The robots moved along paths computed through optimal control synthesis approach to solve complex micromanipulation tasks captured by global LTL formulas. However, the use of local equilibrium points as the states limited the motion of the robot in the workspace to simple tasks. Also, the interaction between the robots limited the robots to stay within far distances with each other. Hence a larger workspace based coil is designed to actuate up to four mm-scale robots in the workspace.<br><br>To improve the resolution of motion of these robots in the workspace, the mm-scale coils are modeled extensively. The forces generated by various coil combinations of the array are modeled and solutions for different actuation force directions are discovered for different locations in the coil. A path planning problem is formulated as a Markov decision process that solves a policy to reach a goal from any location in the workspace. The MDP formulation is also expanded to work when other robots are present in the workspace. The formulation considers the interaction force between the robots and changes the policy to reach the goal location which reduces in the uncertainty of motion of the robot in the presence of interactions from other robots in the workspace.<br><br>The mm-scale coils are difficult to scale down for microrobotic applications and hence a new microscale local magnetic field system was designed. A new microscale local magnetic field system which consisted of two 8 × 8 array of coils aligned in two axes in two layers of a PCB was designed which could actuate robots as small as 1 mm in the plane. The microcoils in the second layer are also able to generate sufficient magnetic field gradients in the workspace, while the traces below it are spaced adequately to eliminate their influence in the workspace. A new microrobot design also enabled the orientation control of the microrobot for performing micromanipulation tasks. However, only two robots could be independently actuated in this workspace due to interaction between the robots.<br><br>In pursuit of actuation smaller and multiple robots in a small workspace, a serpentine coil based local magnetic field generating system was designed to control of the motion of magnets as small as 250 µm. The net size of the robot is 750 µm to enable orientation control and prevent tipping during motion. This system is capable of simultaneous independent closed loop control of up to 4 microrobots. The motion of the robot using the coils resembled that of a stepper motor which enabled the use of sine-cosine functions to specify currents in the coils for smooth motion of the<br>microrobot in the workspace. The experiments demonstrated the capability of the microrobot and platform to simultaneously actuate up to four robots independently and successfully perform manipulation tasks. The ability to control the orientation of the magnet is finally demonstrated that has improved ability to perform manipulation tasks.<br><br>

<div>With the upsurge of greenhouse gas emissions and rapid depletion of fossil fuels, the pressure on the transportation industry to develop new vehicles with improved fuel economy without sacrificing performance is on the rise. Hybrid Electric Vehicles (HEVs), which employ an internal combustion engine as well as an electric motor as power sources, are becoming increasingly popular alternatives to traditional engine only vehicles. However, the presence of multiple power sources makes HEVs more complex. A significant task in developing an HEV is designing a power management strategy, defined as a control system tasked with the responsibility of efficiently splitting the power/torque demand from the separate energy sources. Five different types of power management strategies, which were developed previously, are reviewed in this work, including dynamic programming, equivalent consumption minimization strategy, proportional state-of-charge algorithm, regression modeling and long short term memory modeling. The effects of these power management strategies on the vehicle performance are studied using a simplified model of the vehicle. This work also proposes an original power management strategy development using a genetic algorithm. This power management strategy is compared to dynamic programming and several similarities and differences are observed in the results of dynamic programming and genetic algorithm. For a particular drive cycle, the implementation of the genetic algorithm strategy on the vehicle model leads to a vehicle speed profile that almost matches the original speed profile of that drive cycle.</div>