Inventor
Jeff Stollman
A sampling of recent inventions, not already described, includes
Aerial surveillance system. Designed a surveillance system capable of identifying individual vehicles by vehicle type, speed, and direction. The system met weight restriction to allow easy mounting on small aircraft. Sensor data served as input to pattern recognition software that interpreted the data and identified the vehicles. Vehicle types included motorcycles, cars, pickup trucks, SUVs, and larger trucks.
On line Inspection Method for Pouch Sealing Integrity. Designed a solution to test the seal integrity of pouches containing liquid. The relatively low-tech solution was designed to easily integrate into an existing assembly line. It consisted of having the assembly move pouches through an area with a restricted height. The resulting sandwiching effect applied a fixed amount of pressure to the pouches causing defective pouches to leak. Sensors then detected the leaking pouches and rejected them from the assembly line. The sandwiching mechanism required no automation. Pouches were kept moving by using a rotating belt to form the "ceiling" of the sandwich.
Collision-Avoidance System for Autonomous Consumer Robot. Designed a collision-avoidance system for an autonomous consumer robot, using LIDAR technology. The LIDAR-based vision system detected objects in the path of the robot, calculated their position, and a separate navigation system directed the robot to alter course (if necessary) to move around them. The unit had a 270 degree view capability using a single LIDAR device. This could be increased to 360 degrees using an electro-optic reversible mirror that was deemed unnecessary for the project. Similar vision systems can be applied to some production processes.
Collision-Avoidance System for Autonomous Outdoor Robot. Designed a collision-avoidance system for an autonomous outdoor robot, using LIDAR technology. The LIDAR-based vision system detected objects in the path of the robot, calculated their position, and a separate navigation system directed the robot to alter course (if necessary) to move around them.
Optical Object Recognition and Feature Mapping by Mobile Devices. Managed the development of a mobile application that could identify electronic products by make and model using photos taken by mobile phones. The application could discern make and model even when the equipment was obscured by cables or when the orientation of the photo provided a skewed image.
Visibility Improvement during Bad Weather. Designed a vision system for vehicles based on 3D Flash LIDAR to detect people in conditions of rain, fog, and snow. The system "saw" what was ahead and used various measures of reflectivity to discern live objects from stationery background.
Imaging Technologies for Dark and Stormy Nights. Designed a vision system for use in darkness and bad weather. The system, based on 3D Flash LIDAR uses an active lighting (eye-safe laser pulse) to capture depth/range-detection to further discriminate objects (e.g., people) in the field of view in real-time from background. Each frame capture is instantaneous and measured in nanoseconds. The system can image through +93% obscuration at 40m using an 8.6 degree lens. As well, we have also tested the unit in heavy rain, where we found a degradation of only 10% at 400 meters.
Aerial surveillance system. Designed a surveillance system capable of identifying individual vehicles by vehicle type, speed, and direction. The system met weight restriction to allow easy mounting on small aircraft. Sensor data served as input to pattern recognition software that interpreted the data and identified the vehicles. Vehicle types included motorcycles, cars, pickup trucks, SUVs, and larger trucks.
On line Inspection Method for Pouch Sealing Integrity. Designed a solution to test the seal integrity of pouches containing liquid. The relatively low-tech solution was designed to easily integrate into an existing assembly line. It consisted of having the assembly move pouches through an area with a restricted height. The resulting sandwiching effect applied a fixed amount of pressure to the pouches causing defective pouches to leak. Sensors then detected the leaking pouches and rejected them from the assembly line. The sandwiching mechanism required no automation. Pouches were kept moving by using a rotating belt to form the "ceiling" of the sandwich.
Collision-Avoidance System for Autonomous Consumer Robot. Designed a collision-avoidance system for an autonomous consumer robot, using LIDAR technology. The LIDAR-based vision system detected objects in the path of the robot, calculated their position, and a separate navigation system directed the robot to alter course (if necessary) to move around them. The unit had a 270 degree view capability using a single LIDAR device. This could be increased to 360 degrees using an electro-optic reversible mirror that was deemed unnecessary for the project. Similar vision systems can be applied to some production processes.
Collision-Avoidance System for Autonomous Outdoor Robot. Designed a collision-avoidance system for an autonomous outdoor robot, using LIDAR technology. The LIDAR-based vision system detected objects in the path of the robot, calculated their position, and a separate navigation system directed the robot to alter course (if necessary) to move around them.
Optical Object Recognition and Feature Mapping by Mobile Devices. Managed the development of a mobile application that could identify electronic products by make and model using photos taken by mobile phones. The application could discern make and model even when the equipment was obscured by cables or when the orientation of the photo provided a skewed image.
Visibility Improvement during Bad Weather. Designed a vision system for vehicles based on 3D Flash LIDAR to detect people in conditions of rain, fog, and snow. The system "saw" what was ahead and used various measures of reflectivity to discern live objects from stationery background.
Imaging Technologies for Dark and Stormy Nights. Designed a vision system for use in darkness and bad weather. The system, based on 3D Flash LIDAR uses an active lighting (eye-safe laser pulse) to capture depth/range-detection to further discriminate objects (e.g., people) in the field of view in real-time from background. Each frame capture is instantaneous and measured in nanoseconds. The system can image through +93% obscuration at 40m using an 8.6 degree lens. As well, we have also tested the unit in heavy rain, where we found a degradation of only 10% at 400 meters.