Amazon Technologies, Inc.

Robert Alexander Colburn

James Christopher Curlander

Barry James O'Brien

Andrew Stubbs

Ryan Scott Russell

Gurunandan Krishnan Gorumkonda

Lilian De Greef

Visual task feedback for workstations in a materials handling facility may be implemented. Image data of a workstation surface may be obtained from image sensors. The image data may be evaluated with regard to the performance of an item-handling task at the workstation. The evaluation of the image data may identify items located on the workstation surface, determine a current state of the item-handling task, or recognize an agent gesture at the workstation. Based, at least in part on the evaluation, one or more visual task cues may be selected to project onto the workstation surface. The projection of the selected visual task cues onto the workstation surface may then be directed.

Visual task feedback for workstations in materials handling facilities

The high-intensity specular reflection of light from a surface of an object may create distortions in images of the object, in locations determined based on a perspective of an optical sensor. Multiple images of the object taken from different perspectives may include multiple specular artifacts in different locations, and a composite image of the object that omits such artifacts may be generated based on the multiple images. In particular, pixels of each of the images corresponding to the locations of the specular artifacts and having optimal intensities may be identified. The composite image may be generated based on the pixels having optimal intensities, and by excluding the pixels corresponding to the specular artifacts.

Imaging system for addressing specular reflection

Various examples are directed to systems and methods for utilizing depth videos to analyze material handling tasks. A material handling facility may comprise a depth video system and a control system programmed to receive a plurality of depth videos including performances of the material handling task. For each of the plurality of depth videos, training data may identify sub-tasks of the material handling task and corresponding portions of the video including the sub-tasks. The plurality of depth videos and the training data may be used to train a model to identify the sub-tasks from depth videos. The control system may apply the model to a captured depth video of a human agent performing the material handling task at a workstation to identify a first sub-task of the material handling task being performed by the human agent.

Optimizing material handling tasks

Information regarding actions or activities to be performed at a workstation may be projected upon a portion of the workstation using one or more projectors. The information may include one or more arrows or other indicators referencing specific tools, materials or objects that may be used to perform one or more of the actions or activities. Such arrows or indicators may be rendered in a manner that simulates a three-dimensional or floating appearance thereof from a perspective of a user that may be adjusted or modified based on changes in the perspective of the user, and with respect to one or more physical or virtual sources of light.

Perspective-aware projected user interfaces

An inventory system has inventory pods that are freely and independently moved about a facility and include inventory holders having dynamically reconfigurable storage bins. For various operating scenarios, the components of the inventory system are directed to dynamically reconfigure the storage bins, thereby maintaining efficient product density amongst the inventory holders and permitting the use of automated equipment to manipulate inventory items stored in the inventory pods. One or more inventory pods may be used with lifting modules and picking modules to dynamically reconfigure the storage bins as inventory items are added to and removed from the inventory holders.

Growing community of inventors

Seattle, WA, United States of America

Robert Alexander Colburn