Frederick L. Crabbe

The snackbot: documenting the design of a robot for long-term human-robot interaction

We present the design of the Snackbot, a robot that will deliver snacks in our university buildings. The robot is intended to provide a useful, continuing service and to serve as a research platform for long-term Human-Robot Interaction. Our design process, which occurred over 24 months, is documented as a contribution for others in HRI who may be developing social robots that offer services. We describe the phases of the design project, and the design decisions and tradeoffs that led to the current version of the robot.

SpeechActs: a spoken-language framework

SpeechActs is a prototype testbed for developing spoken natural language applications. In developing SpeechActs, our primary goal was to enable software developers without special expertise in speech or natural language to create effective conversational speech applications-that is, applications with which users can speak naturally, as if they were conversing with a personal assistant. We also wanted SpeechActs applications to work with one another without requiring that each have specific knowledge of other applications running in the same suite. A discourse management component is necessary to embody the information that allows such a natural conversational flow. Because technology changes so rapidly, we also did not want to tie developers to specific speech recognizers or synthesizers. We wanted them to be able to use these speech technologies as plug-in components.

Second-order networks for wall-building agents

Describes robust neurocontrollers for groups of agents that perform construction tasks. They enable agents to balance multiple goals, perform sequences of actions and survive while building walls, corridors, intersections, and briar patches.

Unifying Undergraduate Artificial Intelligence Robotics Layers of Abstraction over Two Channels

From a computer science and artificial intelligence perspective, robotics often appears as a collection of disjoint, sometimes antagonistic subfields. The lack of a coherent and unified presentation of the field negatively affects teaching, especially to undergraduates. This article presents an alternative synthesis of the various subfields of AI robotics and shows how these traditional subfields fit into the whole. Finally, it presents a curriculum based on these ideas.

Design of a low-cost, highly mobile urban search and rescue robot

In this paper, we discuss the design of a novel robotic platform for urban search and rescue. The system developed possesses unique mobility capabilities based on a new adjustable compliance mechanism and overall locomotive morphology. The main facets of this work involve the morphological concepts, initial design and construction of a prototype vehicle, and a physical simulation to be used for developing controllers for semi-autonomous (supervisory) operation.

Compromise strategies for action selection

Among the many properties suggested for action-selection mechanisms, a prominent one is the ability to select compromise actions, i.e. actions that are not the best to satisfy any active goal in isolation, but rather compromise between multiple goals. This paper briefly reviews the history of compromise behaviour and presents experimental analyses of it in an attempt to determine how much compromise behaviour aids an agent. It concludes that optimal compromise behaviour has a surprisingly small benefit over non-compromise behaviour in the experiments performed, and presents some reasons why this may be true and hypothesizes cases where compromise behaviour is truly useful. In particular, it hypothesizes that a crucial factor is the level at which an action is taken (low-level actions are specific, such as ‘move left leg’; high-level actions are vague, such as ‘forage for food’). This paper hypothesizes that compromise behaviour is more beneficial for high- than low-level actions.

Anatomy, dissection, and mechanics of an introductory cyber-security course's curriculum at the United States naval academy

Due to the high priority of cyber-security education, the United States Naval Academy rapidly developed and implemented a new cyber-security course that is required for all of its first-year students. During the fall semester in 2011, half of the incoming class (about 600 students) took the course through a total of 31 sections offered by 16 instructors from a variety of disciplines and backgrounds. In the following spring semester, the remaining half of the first-year students will take the course. This paper explains the motivation that instigated and drove course development, the curriculum, teaching mechanics implemented, personnel required, as well as challenges and lessons learned from the first offering of the course. The information contained in this paper will be useful to those thinking of implementing a technical course required of all students at the same level in an institution (in our case first-year students) and particularly those interested in implementing such a course in cyber security.

On the development of a novel urban search and rescue robot

This paper concerns the development of a novel robotic platform for urban search and rescue (USAR) efforts. The main facets of this work involve the design and construction of a new robot morphology and a physical simulation to be used for developing controllers for semiautonomous (supervisory) operation.

Goal Directed Adaptive Behavior in Second-Order Neural Networks: The MAXSON family of architectures

The paper presents a neural network architecture (MAXSON) based on second-order connections that can learn a multiple goal approach/avoid task using reinforcement from the environment. It also enables an agent to learn vicariously, from the successes and failures of other agents. The paper shows that MAXSON can learn certain spatial navigation tasks much faster than traditional Q-learning, as well as learn goal directed behavior, increasing the agents chances of long-term sur vival. The paper shows that an extension of MAXSON (V-MAXSON) enables agents to learn vicariously, and this improves the overall survivability of the agent population.

Book review: Unifying artificial intelligence robotics: An undergraduate textbook

Robotics as a discipline has been active for more than 40 years and encompasses a wide range of disjoint subfields. Although there is a surfeit of undergraduatelevel textbooks in robotics, history has checked the development of texts appropriate for an undergraduate course from the artificial intelligence (AI) perspective. Robin Murphy’s Introduction to AI Robotics (IAIR) is a new undergraduate text that partially succeeds in remedying this omission. Historically, by far the most developed subfield of robotics is engineering robotics, characterized by a precise mathematical description of the physical properties of the robot (kinematics), and techniques for using this description to execute planned sequences of motions (control theory). The approach is successful for controlling manipulator arms to perform precise, repeatable actions, making them ideal for manufacturing, but engineering robotics never really addresses the problem of enabling the robot to select actions in unforeseen situations. This can be problematic once a robot has a means of locomotion and can get itself into serious trouble. AI robotics developed in parallel to engineering robotics to address this question: How should a mobile robot make decisions for itself? AI robotics is generally regarded to have begun in the 1960s with AI planning techniques still in use today. A robot using these techniques senses the environment, builds a model of the environment, searches for an optimal plan, and then acts based on that plan. Despite numerous difficulties with this approach, it continued to dominate until the 1980s, when several researchers (most visibly Rodney Brooks at MIT) began to point out that we could mitigate many of these difficulties by taking inspiration from biology. The result of this activity was behavior-based (or reactive) robotics, a subfield of AI robotics, where the robot is controlled by a collection of miniature programs, each of which is a tightly coupled mapping from a subset of sensor input to a motor output. The inspirations that led to the establishment of behaviorbased robotics were the same ones that led to the development of the field of adaptive behavior. The genesis of these fields and the revolution centred around them splintered AI robotics with such force that only now is the dust beginning to settle. Unsurprisingly given this volatile history, there has been a lack of an appropriate undergraduate textbook for teaching robotics from the computer science and artificial intelligence perspective. Such a text would need to take into account that few undergraduates can be assumed to have already taken AI courses; allow flexibility in the robotic platform used; and most importantly, it must unify and put in perspective the recently fractured field of robotics. The vast majority of texts in robotics are in engineering robotics. One popular text [Introduction to Robotics by McKerrow, (1991)] spends 431 pages on kinematics, 200 pages on the engineering of sensors and control theory, and only 114 pages on issues connected to mobile intelligent robotics. Some texts, such as Jones and Flynn’s Mobile Robots: Inspiration to Implementation (Jones & Flynn, 1993), cover appropriate AI material but focus on a single robotic platform (in this case the Rug Warrior). Collections of papers, such as Artificial Intelligence and Mobile Robots (Kortenkamp, Bonasso, & Murphy, 1998) can be successful as a text because of their flexibility, but they lack cohesion and are more suitable for Adaptive Behavior