Brassboard development of a MEMS-scanned ladar sensor for small ground robots

摘要

The Army Research Laboratory (ARL) is researching a short-range ladar imager for navigation, obstacle/collision avoidance, and target detection/identification on small unmanned ground vehicles (UGV).To date, commercial UGV ladars have been flawed by one or more factors including low pixelization, insufficient range or range resolution, image artifacts, no daylight operation, large size, high power consumption, and high cost. ARL built a breadboard ladar based on a newly developed but commercially available micro-electro-mechanical system (MEMS) mirror coupled to a lowcost pulsed Erbium fiber laser transmitter that largely addresses these problems. Last year we integrated the ladar and associated control software on an iRobot PackBot and distributed the ladar imagery data via the PackBot's computer network. The un-tethered PackBot was driven through an indoor obstacle course while displaying the ladar data realtime on a remote laptop computer over a wireless link. We later conducted additional driving experiments in cluttered outdoor environments. This year ARL partnered with General Dynamics Robotics Systems to start construction of a brass board ladar design. This paper will discuss refinements and rebuild of the various subsystems including the transmitter and receiver module, the data acquisition and data processing board, and software that will lead to a more compact, lower cost, and better performing ladar. The current ladar breadboard has a 5-6 Hz frame rate, an image size of 256 (h) × 128 (v) pixels, a 60° × 30° field of regard, 20 m range, eyesafe operation, and 40 cm range resolution (with provisions for super-resolution or accuracy).