Amazon Technologies, Inc.

Qicai Shi

Camerin Cole Hahn

Christopher Raymond Grajewski

Nathan Pius O'Neill

Somasundaram Niranjayan

Vinod L Hingorani

Matthew Perkins

Elizabeth Ann Zitzer

A portable device comprises a magnetometer for detecting a magnetic field. When the portable device is at or near a portal, the portable device may detect data from the portal, such as beacon data or a connection message. Responsive to the data from the portal, the portable device operates the magnetometer to detect and characterize a magnetic field generated by the portal, such as by characterizing the magnitude, polarity, and direction of the magnetic field. The portable device generates magnetometer data to indicate the magnetic field strength detected by the magnetometer. The magnetometer data is then transmitted to the portal, along with a device identifier. The portal compares the received magnetometer data to expected magnetic field data to determine presence of the portable device at the portal. The portable device is associated with an account, providing information about entry of a user of that account entering the portal.

System to determine portable device presence at a portal

Sensors at an inventory location can be used to detect the occurrence of an event at the inventory location. For example, capacitive sensors can gather capacitance values, which can then be analyzed for changes that indicate an event has occurred at the inventory location. When an event has been detected, event data is generated and used as a possible trigger to collect additional sensor data. As an example, if an event involving items at an inventory location is detected, such an action can trigger the collection and analysis of additional sensor data, such as weight sensor data, to determine a quantity of the items added or removed from the inventory location.

System to detect event at an inventory location

A facility may include antennas directed to a known location and cameras with a field-of-view that encompasses the known location. The antennas acquire a device identifier and device data from a mobile device carried by a user at the known location. Antenna data is then generated to describe the data acquired by the antennas. Image data is also generated from the cameras to include images of a person at the known location. The antenna data and image data are then processed to determine estimated motion values and a comparison of such values is performed to check if such motion values are within a threshold of one another. If the motion values are within a threshold of each other, data indicative of presence of the mobile device at the known location is generated.

User identification system using directional antennas and cameras

Fixtures can hold one or more items. Load cells may be used to measure the weight of the items on the fixture. In cases where items are added to or removed from the fixture, the load cells will detect a weight change at the fixture. Data reflecting the weight change is then used to determine interactions with the items on the fixture. However, vibrations to the fixture can introduce noise which can lead to inaccuracies in weight data acquired from the load cells. A vibration sensor can be used to measure the level of vibration detected at the fixture. Weight data acquired during a time interval when vibrations exceed a vibration threshold value is deemed invalid weight data. By contrast, weight data acquired when vibrations do not exceed the vibration threshold value will be deemed valid weight data that can be used for determining interactions with items at the fixture.

System to determine effects of vibration on weight data for a fixture

Shelves or other fixtures may be used to support items at a facility. Load cells at the fixtures may be used to acquire weight data indicative of changes to the fixture as items are added or removed from the fixture. Vibration of the fixtures may introduce noise into the weight data, resulting in inaccurate weight data. Described are techniques for utilizing a vibration sensor on the fixture to determine the presence of vibration. In one implementation, weight data obtained during vibration may be disregarded. In another implementation, data from the vibration sensor may be used to adaptively adjust values of the weight data. By utilizing the information from the vibration sensor, more accurate weight data is obtained. The weight data may then be used to determine a count of items at the fixture, determine a quantity of items picked or placed, and so forth.

System to mitigate effects of vibration on load cell

Shelves or other fixtures may be used to hold items at a facility. Load cells at the fixtures may be used to acquire weight data indicative of changes to the fixture as items are added or removed from the fixture. The fixtures, such as a freezer case, may include a door that when opened or closed produces air movement with respect to the shelves. This air movement may produce weight data that could be interpreted as a false event, such as a pick or place, even when none has taken place. A door sensor provides data that is used to determine when the door has been opened or closed. Weight data occurring for some time after the opening of the door or before the closing of the door may be disregarded.

System to process load cell data using door sensor data

Shelves or other fixtures may be used to hold items at a facility. Load cells at the fixtures may be used to acquire weight data indicative of changes to the fixture as items are added or removed from the fixture. The fixtures, such as a freezer case, may include a door that when opened or closed results in air movement with respect to the shelves. This air movement may produce weight data that is erroneously indicative of a false event, such as a pick or place, even when no actual event has taken place. Weight data from load cells in the shelves may be analyzed to determine a correlation in the data from multiple shelves. If there is a high correlation in the changes to the weight data at multiple shelves, the weight data may be disregarded as being due to air movement.

System to process load cell data

Shelves or other fixtures may be used to hold items at a facility. Load cells at the shelf may be used to acquire weight data indicative of changes as items are added or removed from the shelf. The shelves may be part of a fixture that includes a door that results in air movement with respect to the shelves when opened or closed. This air movement may produce weight data that is erroneously indicative of an event, such as a pick or place, even when no actual event has taken place. Weight data from load cells of a shelf may be analyzed to determine a trajectory associated with changes in the weight data over time. A portion of the space may be designated as being associated with invalid weight data. If the trajectory avoids this portion of the space, the weight data may be deemed valid.

System to validate load cell data

An entry portal to a facility includes magnets to produce a magnetic field and a proximity sensor to detect passage of a user. A portable device carried by a user detects the magnetic field and generates magnetometer data which is then sent to the portal along with a device identifier. The magnetometer data is compared to expected magnetic field data to determine presence at the portal. Timestamps of the magnetometer data and data from the proximity sensor may be compared to determine whether the two events occurred contemporaneously. If so, the portable device is deemed to be present at the portal. The device identifier may be used to determine a user, and entry data indicative of entry of the user may be generated. Subsequently, the user may then be tracked using sensors in the facility.

Magnetic entry portal identification system

A user may pick, place, or move an item at an inventory location, such as a shelf. Described are techniques to determine a location of one or more of an object, such as an item or a user, with respect to an array of capacitive sensors. The array may be part of the shelf. As an item is added to, moved or removed from the shelf, when the user's hand is near the shelf, and so forth, capacitance measured by one or more of the sensors in the array may change. Based on these changes, a location relative to the sensor array may be determined. A particular shelf may have different areas, each designated as holding a different type of item. The location information obtained from the capacitive sensors may be used to determine which item on the shelf was interacted with.

Location estimation using array of capacitive sensors

An entry portal to a facility may include radio receivers with directional antennas and cameras at known positions in the facility. The receivers acquire device identifier and device data from a mobile device carried by a user passing through the portal. Portal data is generated that is indicative of the coverage volume of a particular antenna that received a signal carrying the device identifier and signal strength of that signal. The cameras acquire images of the user. Other sensors may acquire other information included in the portal data. The portal data, device data, device identifier, and the images are processed to associate a particular account with a person passing through the entry portal. Subsequently, that person may then be tracked using images from other cameras in the facility.

Entry portal identification system

Various sensors may be used to gather information about items at an inventory location, such as items on a shelf. Weight sensors may be used to gather weight data while capacitive sensors detect the presence of one or more objects such as the items themselves, a user reaching towards the shelf, and so forth. Data from a capacitive sensor may be used to determine that an activity is taking place at the shelf and to trigger processing of weight data from the weight sensor. A combination of a change in capacitance that exceeds a capacitance threshold and a change in weight that exceeds a weight threshold may be used to determine an event, such as a pick or place from the inventory location.

System to determine a change in weight at a load cell

Capacitive sensors may be used to gather information, such as the presence of objects at an inventory location, such as a shelf. Two or more adjacent capacitive sensors operating at the same time may interfere with one another. Described are techniques to prevent this interference by using seed values to initialize pseudo random sequencing of an order in which capacitive sensors are in operation. The capacitive sensor, or a portion thereof, may be also be used as an antenna. Selection of an initial tuning network for use by the antenna may be determined based on output from the capacitive sensor.

Growing community of inventors

Bothell, WA, United States of America

Qicai Shi