*Sunohara & Shibuya Indoor Flight: A Solution for Autonomous Indoor Drone Navigation* * *0:36** Indoor Drone Applications in Japan:* The presentation highlights the increasing demand for indoor drone applications in Japan, including tunnel monitoring, warehouse inventory, factory inspections, and building patrols. * *2:07** Introducing the IND Drone:* Drone Japan introduces the IND drone, featuring ModalAI's VOXL vision system and customizable payloads. * *3:08** Vibration Mitigation:* Efforts were made to reduce vibrations affecting the VOXL and flight controller through the implementation of dampers. * *3:43** Backup Sensor Integration:* A HereFlow optical flow sensor was added as a backup navigation system, providing additional data like velocity. * *4:47** QGroundControl Integration:* The solution utilizes a customized QGroundControl interface, enabling users to create waypoints for automated navigation on a 2D map of the indoor environment. * *5:22** Indoor Flight Procedure:* The workflow involves scaling and uploading a 2D indoor drawing as a map in QGroundControl, assigning internal coordinates for waypoints, and initiating autonomous flight. * *6:36** Demonstration:* A short demonstration showcases the IND drone's navigation capabilities using a 5-meter grid map, with the drone manually moved to simulate flight. * *8:04** VOXL Confidence Monitoring:* A function was added to monitor the quality of ModalAI's VOXL output, displayed as a confidence rate. * *10:06** Technical Advancements:* Randan, a collaborator on the project, discusses key improvements, including: * *10:51** Object Avoidance in Auto Modes:* Implementing a "simple avoidance" feature to stop the drone upon encountering obstacles. * *11:40** Automatic EKF Source Switching:* Enabling seamless switching between VOXL and optical flow data for navigation based on the VOXL confidence level. * *13:54** VOXL Confidence Indication:* Ensuring the VOXL quality metric is visible to the pilot via MAVLink telemetry. * *14:43** Q&A:* The presentation concludes with a Q&A session addressing questions about the VOXL's optical flow capabilities, heat management, EKF source switching smoothness, and performance in dynamic environments. I used gemini-1.5-pro-exp-0827 on rocketrecap dot com to summarize the transcript. Cost (if I didn't use the free tier): $0.02 Input tokens: 16131 Output tokens: 487
Great advance in the VOXL integration. These are the features I wished for during my development ;-) If anyone interested I have a brand new VOXL CAM for sale at half price , Just PM me for details
*Sunohara & Shibuya Indoor Flight: A Solution for Autonomous Indoor Drone Navigation*
* *0:36** Indoor Drone Applications in Japan:* The presentation highlights the increasing demand for indoor drone applications in Japan, including tunnel monitoring, warehouse inventory, factory inspections, and building patrols.
* *2:07** Introducing the IND Drone:* Drone Japan introduces the IND drone, featuring ModalAI's VOXL vision system and customizable payloads.
* *3:08** Vibration Mitigation:* Efforts were made to reduce vibrations affecting the VOXL and flight controller through the implementation of dampers.
* *3:43** Backup Sensor Integration:* A HereFlow optical flow sensor was added as a backup navigation system, providing additional data like velocity.
* *4:47** QGroundControl Integration:* The solution utilizes a customized QGroundControl interface, enabling users to create waypoints for automated navigation on a 2D map of the indoor environment.
* *5:22** Indoor Flight Procedure:* The workflow involves scaling and uploading a 2D indoor drawing as a map in QGroundControl, assigning internal coordinates for waypoints, and initiating autonomous flight.
* *6:36** Demonstration:* A short demonstration showcases the IND drone's navigation capabilities using a 5-meter grid map, with the drone manually moved to simulate flight.
* *8:04** VOXL Confidence Monitoring:* A function was added to monitor the quality of ModalAI's VOXL output, displayed as a confidence rate.
* *10:06** Technical Advancements:* Randan, a collaborator on the project, discusses key improvements, including:
* *10:51** Object Avoidance in Auto Modes:* Implementing a "simple avoidance" feature to stop the drone upon encountering obstacles.
* *11:40** Automatic EKF Source Switching:* Enabling seamless switching between VOXL and optical flow data for navigation based on the VOXL confidence level.
* *13:54** VOXL Confidence Indication:* Ensuring the VOXL quality metric is visible to the pilot via MAVLink telemetry.
* *14:43** Q&A:* The presentation concludes with a Q&A session addressing questions about the VOXL's optical flow capabilities, heat management, EKF source switching smoothness, and performance in dynamic environments.
I used gemini-1.5-pro-exp-0827 on rocketrecap dot com to summarize the transcript.
Cost (if I didn't use the free tier): $0.02
Input tokens: 16131
Output tokens: 487
Great advance in the VOXL integration. These are the features I wished for during my development ;-)
If anyone interested I have a brand new VOXL CAM for sale at half price , Just PM me for details