Knowledge-based interpretation of architectural drawings

A method for machine interpretation of architectural (or other) schematic drawings is presented. The central problem is to build an efficient drawing parser (i.e., a program that identifies the semantic entities, characteristics, and relationships that are represented in the drawing). The parser is built from specifications of the drawing grammar and an underlying spatial model. The grammar describes what to look for, and the spatial model enables the parser to find it quickly. Coupled with existing optical recognition technology, this technique enables the use of drawings directly as: (1) a database to drive various Architecture-Engineering-Construction (AEC) applications; (2) a communication protocol to integrate CAD systems; (3) a traditional user interface.     

Indoor robot gardening: design and implementation

This paper describes the architecture and implementation of a distributed autonomous gardening system with applications in urban/indoor precision agriculture. The garden is a mesh network of robots and plants. The gardening robots are mobile manipulators with an eye-in-hand camera. They are capable of locating plants in the garden, watering them, and locating and grasping fruit. The plants are potted cherry tomatoes enhanced with sensors and computation to monitor their well-being (e.g. soil humidity, state of fruits) and with networking to communicate servicing requests to the robots. By embedding sensing, computation, and communication into the pots, task allocation in the system is de-centrally coordinated, which makes the system scalable and robust against the failure of a centralized agent. We describe the architecture of this system and present experimental results for navigation, object recognition, and manipulation as well as challenges that lie ahead toward autonomous precision agriculture with multi-robot teams.     

Experiments on Surface Reconstruction for Partially Submerged Marine Structures

Over the past 10 years, significant scientific effort has been dedicated to the problem of three‐dimensional (3‐D) surface reconstruction for structural systems. However, the critical area of marine structures remains insufficiently studied. The research presented here focuses on the problem of 3‐D surface reconstruction in the marine environment. This paper summarizes our hardware, software, and experimental contributions on surface reconstruction over the past few years (2008–2011). We propose the use of off‐the‐shelf sensors and a robotic platform to scan marine structures both above and below the waterline, and we develop a method and software system that uses the Ball Pivoting Algorithm (BPA) and the Poisson reconstruction algorithm to reconstruct 3‐D surface models of marine structures from the scanned data. We have tested our hardware and software systems extensively in Singapore waters, including operating in rough waters, where water currents are around 1–2 m/s. We present results on construction of various 3‐D models of marine structures, including slowly moving structures such as floating platforms, moving boats, and stationary jetties. Furthermore, the proposed surface reconstruction algorithm makes no use of any navigation sensor such as GPS, a Doppler velocity log, or an inertial navigation system.     

Congenital infection by human parvovirus B19 ascites-anaemia

A case of intrauterine infection by human parvovirus B19 (HPV B19) manifested as ascites during pregnancy is presented. Ascites was diagnosed by ultrasound at 27 weeks' gestation. A caesarean section was performed at 37 weeks'. owing to affected mobility of the fetus. A pale, female infant with low haemoglobin and bradycardia was delivered. Polymerace Chain Reaction (PCR) lab tests revealed that the mother and the fetus were infected by HPV B19. The neonate was born with low haemoglobin (Hb = 10 g/dl) and with ascites; it was discharged in good general condition 50 days after delivery.     

An automatic coarse and fine surface mesh generation scheme based on medial axis transform: Part II implementation

In this paper, we present implementation aspects of a surface finite element (FE) meshing algorithm described in Part I (this volume) [1]. This meshing scheme is based on the medial axis transform (MAT) [2] to interrogate shape and to subdivide it into topologically simple subdomains. The algorithm can be effectively used to create coarse discretization and fine triangular surface meshes. We describe our techniques and methodology used in the implementation of the meshing and MAT algorithms. We also present some running times of our experimental system. We finally report the results we have obtained from several design and analysis applications which include adaptive surface approximations using triangular facets, and adaptiveh- andp-adaptive finite element analysis (FEA) of plane stress problems. These studies demonstrate the potential applicability of our techniques in computer aided design and analysis.     

Representation of piecewise continuous algebraic surfaces in terms of B-splines

A method for representing shape using portions of algebraic surfaces bounded by rectangular boxes defined in terms of triple product Bernstein polynomials is described and some of its properties are outlined. The method is extended to handle piecewise continuous algebraic surfaces within rectangular boxes defined in terms of triple products of B-spline basis functions. Next, two techniques for sculptured shape creation are studied. The first is based on geometric manipulation of existing primitives and the second on approximation/interpolation of lower dimensional entities using least-squares techniques based on singular value decomposition. In addition, several interrogation techniques, such as contouring, ray tracing and curvature evaluation, used in the design and analysis of piecewise continuous algebraic surfaces are discussed.     

Solving nonlinear polynomial systems in the barycentric Bernstein basis

We present a method for solving arbitrary systems of N nonlinear polynomials in n variables over an n-dimensional simplicial domain based on polynomial representation in the barycentric Bernstein basis and subdivision. The roots are approximated to arbitrary precision by iteratively constructing a series of smaller bounding simplices. We use geometric subdivision to isolate multiple roots within a simplex. An algorithm implementing this method in rounded interval arithmetic is described and analyzed. We find that when the total order of polynomials is close to the maximum order of each variable, an iteration of this solver algorithm is asymptotically more efficient than the corresponding step in a similar algorithm which relies on polynomial representation in the tensor product Bernstein basis. We also discuss various implementation issues and identify topics for further study.