<p>Visual navigation of mobile robots has become a core capability that enables many interesting applications from planetary exploration to self-driving cars. While systems built on passive cameras have been shown to be robust in well-lit scenes, they cannot handle the range of conditions associated with a full diurnal cycle. Lidar, which is fairly invariant to ambient lighting conditions, offers one possible remedy to this problem. In this paper, we describe a visual navigation pipeline that exploits lidar’s ability to measure both range and intensity (a.k.a., reflectance) information. In particular, we use <em>lidar intensity images</em> (from a scanning-laser rangefinder) to carry out tasks such as <em>visual odometry</em> (VO) and <em>visual teach and repeat</em> (VT&R) in realtime, from full-light to full-dark conditions. This lighting invariance comes at the price of coping with motion distortion, owing to the scanning-while-moving nature of laser-based imagers. We present our results and lessons learned from the last few years of research in this area.</p>
We have developed techniques for Simultaneous Localization and Map Building based on the augmented state Kalman filter, and demonstrated this in real time using laboratory robots. Here we report the results of experiments conducted out doors in an unstructured, unknown, representative environment, using a van equipped with a laser range finder for sensing the external environment, and GPS to provide an estimate of ground truth. The goal is simultaneously to build a map of an unknown environment and to use that map to navigate a vehicle that otherwise would have no way of knowing its location. In this paper we describe the system architecture, the nature of the experimental set up, and the results obtained. These are compared with the estimated ground truth. We show that SLAM is both feasible and useful in real environments. In particular, we explore its repeatability and accuracy, and discuss some practical implementation issues. Finally, we look at the way forward for a real implementation on ground and air vehicles operating in very demanding, harsh environments.
Fire and light have long symbolized the relationship of human beings to the universe and its creators. In South and Southeast Asia and the Himalayas, the lamp, as a bearer of light, came to be perceived as a vehicle through which the divine could be accessed. The design, construction, and use of the lamp in these regions have been synonymous with the faith of the devotee since ancient times. Today, the lamp continues to play a pivotal role in Hindu and Buddhist religious contexts, allowing the faithful to concentrate on the image or nature of the deity. The 76 remarkable metal lamps and incense burners illustrated in Flames of Devotion form the heart of a collection assembled by the preeminent scholar of Indian and Himalayan art Pratapaditya Pal and his wife, Chitralekha. They are noteworthy for their ingenious design and diverse crafting, as well as their iconographic richness. They represent fourteen states in India, a majority coming from Rajasthan and Gujarat in the west, the tribal areas in central Madhya Pradesh and Chhattisgarh, and the southern states of Kerala and Tamil Nadu. In addition, stunning examples of lamps from Nepal and Tibet showcase the skill with which precious metals were employed during the eighteenth and nineteenth centuries, and a small selection of early incense burners and lamps from Cambodia, Indonesia, and Vietnam show the role these objects played in the ancient imagination. In an engaging and highly informative text, architect and art historian Sean Anderson investigates why lamps have endured and remain omnipresent in Hindu and Buddhist practice. While examining the historical importance of the lamp, Anderson emphasizes that as altar and tool, icon and fine sculpture, it is an evocative reminder of an undying devotion forged with the most common yet enigmatic of materials: metal and fire. He considers as well the liminal space the lamp occupies between the secular and the sacred in societies where it is often used to mark every event of significance from birth to death.
There is substantial evidence that repeating cerebellar microcircuit implements an adaptive filter algorithm suitable for fine tuning a wide range of sensorimotor skills [1]. Although it is known that multimodal map layer of superior colliculus receives a massive cerebellar input which directly influences collicular output cells [2], there has been little speculation as to function of this influence. We suggest here that these cerebellar inputs play a role in calibrating accuracy of collicular topographic maps.
The cerebellum is a plausible candidate for this role for a number of reasons. It is known to be crucial for accuracy of saccades generated by colliculus. It is also known that cerebellar lesions impair prism adaptation of eye movements. A recent imaging study [3] found direct evidence of cerebellar activation during prism adaptation and suggests that the cerebellum is particularly involved in [establishing] a correct spatial mapping among visuomotor and sensorimotor coordinates systems. A modeling study [3] of collicular visuomotor mapping concludes that transformation occurs in brainstem with cerebellum adjusting it for accuracy, stating that ... importance of cerebellum has been neglected in previous modeling studies.
Here we investigate whether adaptive filter model of cerebellar microcircuit which has been so successful in conventional sensorimotor contexts can be applied without change to very different computational problem of calibrating a topographic map driving an orienting response. We propose a model in which (i) unimodal sensory topographic maps constitute a probabilistic representation of target position and that these maps are optimally combined to produce a multimodal map [4] whose peak activity drives orienting response, (ii) cerebellar input can bias position of peak map activity (e.g. by a process such as attentional gain modulation [5]), (iii) cerebellum receives sensory information that generates topographic maps on its mossy fibre inputs, and (iv) information about orienting errors caused by miscalibration is made available to cerebellum on its climbing fibre inputs and drives cerebellar learning.
We demonstrate in simulation that this mechanism can successfully calibrate topographic maps and go on to investigate its computational properties. For example we investigate a fundamental calibration ambiguity in which sensory maps can be miscalibrated in such a way that their effects on combined estimate cancel. We show that this ambiguity is resolved in a plausible way if error signals are gated whenever a sensor fails to observe target; a similar gating process has been observed in some motor behaviors [6].We also demonstrate that optimality of cerebellar learning rule [7] ensures that Purkinje cell synapses carrying cross-talk between sensors are driven to silence, greatly reducing need to hard-wire connectivity of parallel fiber inputs to cerebellar microzones.
Avra Valley is a north-trending alluvial basin about 15 mi west of Tucson in Pima and Pinal Counties in south-central Arizona. The valley includes about 520 sq mi of which about 100 sq mi is in the San Xavier Indian Reservation. The basin is bounded on the east by the Tortolita, Tucson, and Sierrita Mountains and on the west by the Picacho, Silverbell, and Roskruge Mountains. The climate of the valley is semiarid, the average annual precipitation ranges from 8 to 12 in., and the average annual lake evaporation ranges from 58 to 62 in. Two major ephemeral streams--Santa Cruz River and Brawley Wash--drain the area. Santa Cruz River and Brawley Wash and their tributaries provide a source of recharge to an extensive alluvial aquifer that underlies the valley floor. Since 1940, the amount of groundwater pumped from the aquifer has been greater than the amount of natural recharge from infiltration and underflow. Overdraft of the aquifer resulted in substantial water level declines throughout the valley. Until 1969, use of groundwater in Avra Valley was for irrigation. Since 1969, the city of Tucson has pumped and transported groundwater for municipal use in the adjacent Tucson basin from lands that were purchased and retired from agriculture. The purpose of this report is to describe groundwater conditions in Avra Valley as of 1985. A brief discussion of the geohydrologic setting and history of groundwater development are given to define aquifer characteristics, changes in groundwater levels, and groundwater pumpage since 1940. (Lantz-PTT)
The pollen grains of several gymnosperm groups consist of a main body and one to three air-filled bladders, or sacci. Although sacci may serve a buoyancy function to orient the grain on the ovular pollination droplet in some taxa, sacci have also been shown to increase pollen volume while adding minimal mass, thus decreasing density and thereby increasing the aerodynamic efficiency of wind pollination. However, no published studies have quantitatively addressed the effects of grain geometry or surface ornamentation at the low Reynolds numbers that pollen grains demonstrate. The objectives of this study were to empirically investigate the effects of varying geometries and surface ornamentation on the aerodynamic properties of saccate pollen grains through the experimental determination of drag coefficients and shape factors. Structurally different grains of two extant conifers (Pinus and Falcatifolium) were studied, and using electron microscopy, mathematical modeling, and solid modeling, we created scaled-up physical models of the pollen types. Models were produced with and without sacci, as well as with and without surface texture on the main body. Sacci increased the shape factor, or resistance coefficient, in all pollen types studied, compared to the same types that had been modeled without sacci. The presence of surface ornamentation also decreased the drag coefficients for saccate pollen grains of Pinus. This study is the first to experimentally demonstrate the effect of surface texture on drag for any biological or nonbiological particle at low Reynolds numbers. This study also provides additional empirical evidence for the aerodynamic role of sacci, supporting their adaptive significance for anemophily.
Appearance-based batch nonlinear optimization techniques for simultaneous localization and mapping (SLAM) have been highly successful in assisting robot motion estimation. Traditionally, these techniques are applied in a single privileged coordinate frame, which can become computationally expensive over long distances, particularly when a loop closure requires the adjustment of many pose variables. Recent approaches to the problem have shown that a completely relative coordinate framework can be used to incrementally find a close approximation of the full maximum likelihood solution in constant time. However, due to the nature of these discrete-time techniques, the state size becomes intractable when challenged with high-rate sensors. We propose moving the relative coordinate formulation of SLAM into continuous time by estimating the velocity profile of the robot. We derive the relative formulation of the continuous-time robot trajectory and formulate an estimator for the SLAM problem using temporal basis functions. Although we do not yet take advantage of large-scale loop closures, we intentionally use a relative formulation to set the stage for future work that will close loops in constant time. We show how the estimator can be used in a window-style filter to incrementally find the batch solution in constant time. The estimator is validated on a set of appearance-based feature measurements acquired using a two-axis scanning laser rangefinder over a 1.1km trajectory.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
