Matching Items (26)
Filtering by
- All Subjects: Robotics
- Creators: Computer Science and Engineering Program
Description
Robots are often used in long-duration scenarios, such as on the surface of Mars,where they may need to adapt to environmental changes. Typically, robots have been built specifically for single tasks, such as moving boxes in a warehouse or surveying construction sites. However, there is a modern trend away from human hand-engineering and toward robot learning. To this end, the ideal robot is not engineered,but automatically designed for a specific task. This thesis focuses on robots which learn path-planning algorithms for specific environments. Learning is accomplished via genetic programming. Path-planners are represented as Python code, which is optimized via Pareto evolution. These planners are encouraged to explore curiously and efficiently. This research asks the questions: “How can robots exhibit life-long learning where they adapt to changing environments in a robust way?”, and “How can robots learn to be curious?”.
ContributorsSaldyt, Lucas P (Author) / Ben Amor, Heni (Thesis director) / Pavlic, Theodore (Committee member) / Computer Science and Engineering Program (Contributor, Contributor) / Barrett, The Honors College (Contributor)
Created2021-05
Description
Not enough students are earning bachelor’s degrees in Computer Science, which is shocking as computing jobs are growing by the thousands (Zampa, 2016). These jobs have high-paying salaries and are not going to fade from the future any time soon, that is why the falling rates of computer science graduates are alarming. The working hypothesis on why so few college students major in computer science is that most think that it is too hard to learn (Wang, 2017). But I believe the real reason lies in that computer science is not an educational subject that is taught before university, which is too late for most students because by ages 12 to 13 (about seventh to eighth grade) they have decided that computer science concepts are “too difficult” for them to learn (Learning, 2022). Implementing a computer science-based education at an earlier age can possibly circumvent this seen development where students begin to lose confidence and doubt their abilities to learn computer science. This can be done easily by integrating computer science into academic subjects that are already taught in elementary schools such as science, math, and language arts as computer science uses logic, syntax, and other skills that are broadly applicable. Thus, I have created a introductory lesson plan for an elementary school class that incorporates learning how to code with robotics to promote learning computer science principles and destigmatize that it is “too hard” to learn in university.
ContributorsWong, Erika (Author) / Hedges, Craig (Thesis director) / Fischer, Adelheid (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2023-05
Description
Preventive maintenance is a practice that has become popular in recent years, largely due to the increased dependency on electronics and other mechanical systems in modern technologies. The main idea of preventive maintenance is to take care of maintenance-type issues before they fully appear or cause disruption of processes and daily operations. One of the most important parts is being able to predict and foreshadow failures in the system, in order to make sure that those are fixed before they turn into large issues. One specific area where preventive maintenance is a very big part of daily activity is the automotive industry. Automobile owners are encouraged to take their cars in for maintenance on a routine schedule (based on mileage or time), or when their car signals that there is an issue (low oil levels for example). Although this level of maintenance is enough when people are in charge of cars, the rise of autonomous vehicles, specifically self-driving cars, changes that. Now instead of a human being able to look at a car and diagnose any issues, the car needs to be able to do this itself. The objective of this project was to create such a system. The Electronics Preventive Maintenance System is an internal system that is designed to meet all these criteria and more. The EPMS system is comprised of a central computer which monitors all major electronic components in an autonomous vehicle through the use of standard off-the-shelf sensors. The central computer compiles the sensor data, and is able to sort and analyze the readings. The filtered data is run through several mathematical models, each of which diagnoses issues in different parts of the vehicle. The data for each component in the vehicle is compared to pre-set operating conditions. These operating conditions are set in order to encompass all normal ranges of output. If the sensor data is outside the margins, the warning and deviation are recorded and a severity level is calculated. In addition to the individual focus, there's also a vehicle-wide model, which predicts how necessary maintenance is for the vehicle. All of these results are analyzed by a simple heuristic algorithm and a decision is made for the vehicle's health status, which is sent out to the Fleet Management System. This system allows for accurate, effortless monitoring of all parts of an autonomous vehicle as well as predictive modeling that allows the system to determine maintenance needs. With this system, human inspectors are no longer necessary for a fleet of autonomous vehicles. Instead, the Fleet Management System is able to oversee inspections, and the system operator is able to set parameters to decide when to send cars for maintenance. All the models used for the sensor and component analysis are tailored specifically to the vehicle. The models and operating margins are created using empirical data collected during normal testing operations. The system is modular and can be used in a variety of different vehicle platforms, including underwater autonomous vehicles and aerial vehicles.
ContributorsMian, Sami T. (Author) / Collofello, James (Thesis director) / Chen, Yinong (Committee member) / School of Mathematical and Statistical Sciences (Contributor) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05
Description
Education in computer science is a difficult endeavor, with learning a new programing language being a barrier to entry, especially for college freshman and high school students. Learning a first programming language requires understanding the syntax of the language, the algorithms to use, and any additional complexities the language carries. Often times this becomes a deterrent from learning computer science at all. Especially in high school, students may not want to spend a year or more simply learning the syntax of a programming language. In order to overcome these issues, as well as to mitigate the issues caused by Microsoft discontinuing their Visual Programming Language (VPL), we have decided to implement a new VPL, ASU-VPL, based on Microsoft's VPL. ASU-VPL provides an environment where users can focus on algorithms and worry less about syntactic issues. ASU-VPL was built with the concepts of Robot as a Service and workflow based development in mind. As such, ASU-VPL is designed with the intention of allowing web services to be added to the toolbox (e.g. WSDL and REST services). ASU-VPL has strong support for multithreaded operations, including event driven development, and is built with Microsoft VPL users in mind. It provides support for many different robots, including Lego's third generation robots, i.e. EV3, and any open platform robots. To demonstrate the capabilities of ASU-VPL, this paper details the creation of an Intel Edison based robot and the use of ASU-VPL for programming both the Intel based robot and an EV3 robot. This paper will also discuss differences between ASU-VPL and Microsoft VPL as well as differences between developing for the EV3 and for an open platform robot.
ContributorsDe Luca, Gennaro (Author) / Chen, Yinong (Thesis director) / Cheng, Calvin (Committee member) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2015-12
Description
Many real-world planning problems can be modeled as Markov Decision Processes (MDPs) which provide a framework for handling uncertainty in outcomes of action executions. A solution to such a planning problem is a policy that handles possible contingencies that could arise during execution. MDP solvers typically construct policies for a problem instance without re-using information from previously solved instances. Research in generalized planning has demonstrated the utility of constructing algorithm-like plans that reuse such information. However, using such techniques in an MDP setting has not been adequately explored.
This thesis presents a novel approach for learning generalized partial policies that can be used to solve problems with different object names and/or object quantities using very few example policies for learning. This approach uses abstraction for state representation, which allows the identification of patterns in solutions such as loops that are agnostic to problem-specific properties. This thesis also presents some theoretical results related to the uniqueness and succinctness of the policies computed using such a representation. The presented algorithm can be used as fast, yet greedy and incomplete method for policy computation while falling back to a complete policy search algorithm when needed. Extensive empirical evaluation on discrete MDP benchmarks shows that this approach generalizes effectively and is often able to solve problems much faster than existing state-of-art discrete MDP solvers. Finally, the practical applicability of this approach is demonstrated by incorporating it in an anytime stochastic task and motion planning framework to successfully construct free-standing tower structures using Keva planks.
This thesis presents a novel approach for learning generalized partial policies that can be used to solve problems with different object names and/or object quantities using very few example policies for learning. This approach uses abstraction for state representation, which allows the identification of patterns in solutions such as loops that are agnostic to problem-specific properties. This thesis also presents some theoretical results related to the uniqueness and succinctness of the policies computed using such a representation. The presented algorithm can be used as fast, yet greedy and incomplete method for policy computation while falling back to a complete policy search algorithm when needed. Extensive empirical evaluation on discrete MDP benchmarks shows that this approach generalizes effectively and is often able to solve problems much faster than existing state-of-art discrete MDP solvers. Finally, the practical applicability of this approach is demonstrated by incorporating it in an anytime stochastic task and motion planning framework to successfully construct free-standing tower structures using Keva planks.
ContributorsKala Vasudevan, Deepak (Author) / Srivastava, Siddharth (Thesis advisor) / Zhang, Yu (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2020
Description
Most planning agents assume complete knowledge of the domain, which may not be the case in scenarios where certain domain knowledge is missing. This problem could be due to design flaws or arise from domain ramifications or qualifications. In such cases, planning algorithms could produce highly undesirable behaviors. Planning with incomplete domain knowledge is more challenging than partial observability in the sense that the planning agent is unaware of the existence of such knowledge, in contrast to it being just unobservable or partially observable. That is the difference between known unknowns and unknown unknowns.
In this thesis, I introduce and formulate this as the problem of Domain Concretization, which is inverse to domain abstraction studied extensively before. Furthermore, I present a solution that starts from the incomplete domain model provided to the agent by the designer and uses teacher traces from human users to determine the candidate model set under a minimalistic model assumption. A robust plan is then generated for the maximum probability of success under the set of candidate models. In addition to a standard search formulation in the model-space, I propose a sample-based search method and also an online version of it to improve search time. The solution presented has been evaluated on various International Planning Competition domains where incompleteness was introduced by deleting certain predicates from the complete domain model. The solution is also tested in a robot simulation domain to illustrate its effectiveness in handling incomplete domain knowledge. The results show that the plan generated by the algorithm increases the plan success rate without impacting action cost too much.
In this thesis, I introduce and formulate this as the problem of Domain Concretization, which is inverse to domain abstraction studied extensively before. Furthermore, I present a solution that starts from the incomplete domain model provided to the agent by the designer and uses teacher traces from human users to determine the candidate model set under a minimalistic model assumption. A robust plan is then generated for the maximum probability of success under the set of candidate models. In addition to a standard search formulation in the model-space, I propose a sample-based search method and also an online version of it to improve search time. The solution presented has been evaluated on various International Planning Competition domains where incompleteness was introduced by deleting certain predicates from the complete domain model. The solution is also tested in a robot simulation domain to illustrate its effectiveness in handling incomplete domain knowledge. The results show that the plan generated by the algorithm increases the plan success rate without impacting action cost too much.
ContributorsSharma, Akshay (Author) / Zhang, Yu (Thesis advisor) / Fainekos, Georgios (Committee member) / Srivastava, Siddharth (Committee member) / Arizona State University (Publisher)
Created2020
Description
Autonomous systems powered by Artificial Neural Networks (NNs) have shown remarkable capabilities in performing complex tasks that are difficult to formally specify. However, ensuring the safety, reliability, and trustworthiness of these NN-based systems remains a significant challenge, especially when they encounter inputs that fall outside the distribution of their training data. In robot learning applications, such as lower-leg prostheses, even well-trained policies can exhibit unsafe behaviors when faced with unforeseen or adversarial inputs, potentially leading to harmful outcomes. Addressing these safety concerns is crucial for the adoption and deployment of autonomous systems in real-world, safety-critical environments. To address these challenges, this dissertation presents a neural network repair framework aimed at enhancing safety in robot learning applications. First, a novel layer-wise repair method utilizing Mixed-Integer Quadratic Programming (MIQP) is introduced that enables targeted adjustments to specific layers of a neural network to satisfy predefined safety constraints without altering the network’s structure. Second, the practical effectiveness of the proposed methods is demonstrated through extensive experiments on safety-critical assistive devices, particularly lower-leg prostheses, to ensure the generation of safe and reliable neural policies. Third, the integration of predictive models is explored to enforce implicit safety constraints, allowing for anticipation and mitigation of unsafe behaviors through a two-step supervised learning approach that combines behavioral cloning with neural network repair. By addressing these areas, this dissertation advances the state-of-the-art in neural network repair for robot learning. The outcome of this work promotes the development of robust and secure autonomous systems capable of operating safely in unpredictable and dynamic real-world environments.
ContributorsMajd, Keyvan (Author) / Ben Amor, Heni (Thesis advisor) / Fainekos, Georgios (Thesis advisor) / Srivastava, Siddharth (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2024
Description
Reinforcement learning (RL) is a powerful methodology for teaching autonomous agents complex behaviors and skills. A critical component in most RL algorithms is the reward function -- a mathematical function that provides numerical estimates for desirable and undesirable states. Typically, the reward function must be hand-designed by a human expert and, as a result, the scope of a robot's autonomy and ability to safely explore and learn in new and unforeseen environments is constrained by the specifics of the designed reward function. In this thesis, I design and implement a stateful collision anticipation model with powerful predictive capability based upon my research of sequential data modeling and modern recurrent neural networks. I also develop deep reinforcement learning methods whose rewards are generated by self-supervised training and intrinsic signals. The main objective is to work towards the development of resilient robots that can learn to anticipate and avoid damaging interactions by combining visual and proprioceptive cues from internal sensors. The introduced solutions are inspired by pain pathways in humans and animals, because such pathways are known to guide decision-making processes and promote self-preservation. A new "robot dodge ball' benchmark is introduced in order to test the validity of the developed algorithms in dynamic environments.
ContributorsRichardson, Trevor W (Author) / Ben Amor, Heni (Thesis advisor) / Yang, Yezhou (Committee member) / Srivastava, Siddharth (Committee member) / Arizona State University (Publisher)
Created2018
Description
In this thesis, a new approach to learning-based planning is presented where critical regions of an environment with low probability measure are learned from a given set of motion plans. Critical regions are learned using convolutional neural networks (CNN) to improve sampling processes for motion planning (MP).
In addition to an identification network, a new sampling-based motion planner, Learn and Link, is introduced. This planner leverages critical regions to overcome the limitations of uniform sampling while still maintaining guarantees of correctness inherent to sampling-based algorithms. Learn and Link is evaluated against planners from the Open Motion Planning Library (OMPL) on an extensive suite of challenging navigation planning problems. This work shows that critical areas of an environment are learnable, and can be used by Learn and Link to solve MP problems with far less planning time than existing sampling-based planners.
In addition to an identification network, a new sampling-based motion planner, Learn and Link, is introduced. This planner leverages critical regions to overcome the limitations of uniform sampling while still maintaining guarantees of correctness inherent to sampling-based algorithms. Learn and Link is evaluated against planners from the Open Motion Planning Library (OMPL) on an extensive suite of challenging navigation planning problems. This work shows that critical areas of an environment are learnable, and can be used by Learn and Link to solve MP problems with far less planning time than existing sampling-based planners.
ContributorsMolina, Daniel, M.S (Author) / Srivastava, Siddharth (Thesis advisor) / Li, Baoxin (Committee member) / Zhang, Yu (Committee member) / Arizona State University (Publisher)
Created2019
Description
Robots are increasingly integrated into our daily routines, yet their current tasks are mostly specific and pre-programmed, lacking flexibility and scalability. In anticipation of a future where robots will handle diverse household chores—from basic tasks like picking and placing items to more complex activities such as cooking—there's a critical need for them to master long-term planning and motion challenges. Current methods addressing this demand typically rely on manually crafted abstractions and expert-guided task planning. This work deals with a novel approach - developing strategies to learn relational abstractions directly from raw trajectory data. These learned abstractions are then used for inventing symbolic vocabularies and action models. The learnt action models are then used to solve complex long-horizon task and motion planning problems, which are not seen in the training demonstrations. The results show that the approach discussed is robust and capable of learning the model just from few demonstrations. Additionally, this work also discusses an interactive AI platform aimed at making advanced robot planning accessible to users without extensive computer science backgrounds. Such platforms play a crucial role as AI and robotics increasingly intertwine with everyday life, offering intuitive interfaces that teach users the basics of robot planning.
ContributorsNagpal, Jayesh (Author) / Srivastava, Siddharth (Thesis advisor) / Pedrielli, Giulia (Committee member) / Gopalan, Nakul (Committee member) / Arizona State University (Publisher)
Created2024