PAWAN S RAO's Posts (1)

Sort by

PIXHAWK ON CUSTOM BUILT AIR-FRAME - 'DAWON'

3689637624?profile=original

We are Team Edhitha, a team of undergrads from M S Ramaiah Institute of Technology Bangalore India. We have been participating in Student Unmanned Aerial Systems (SUAS) organized by Association For Unmanned Vehicles Systems International (AUVSI) from past 4 years. The competition requires the team to design and deploy a UAS with capabilities of Auto Navigation with takeoff and landing,  Auto Target detection by Image Processing, Air Drop, IR imagery, Interoperability etc.

Last year 3D robotics sponsored a Pixhawk auto-pilot kit and we would like to express our gratitude for extending help. We are pleased to inform you that the team has done very well in the SUAS 2014 competition (June 18-22 2014). Using the Pixhawk kit, we put up a stunning display of automated flight in the competition winning 7th position in the oral Flight readiness review (FRR) and 10th place in Journal Paper Presentation and 21st position in Mission among 49 International teams !

It has been a great learning time with an exponential learning curve with the Pixhawk with many ups and downs which I'd like to share here. Here is our journey with Pixhawk ....

Firstly to understand the new AP we integrated the new Pixhawk on 3 platforms - Electric, Nitro and Gas. We faced quite many deadsticks on nitro and gas planes, so we learnt how to tune AP by throttle setting like slew rates, safe idle adjustment etc. Now with an experience of having worked on Gas and Nitro platforms we got a confidence to incorporate the AP on our Custom Built Plane DAWON ! (2.6m wingspan, twin boom with A-tail configuration powered by a 30 cc gas engine ). A lot of effort has been put into perfecting the operation of Pixhawk autopilot on our indigenous  developed airframe. (Here is an inspiring story behind inception of DAWON project, don't miss this - http://www.edhithadrone.com/dawon-project.html)

 

In Dawon however we were faced with a particularly problematic situation when all 7 digital servos were connected to the Pixhawk. During flights there were frequent blackouts of the autopilot wherein the entire system would go into reboot while the servos simply went limp. Having lost control over the aircraft we even lost one of our airframes in a crash!.  In subsequent tests the problem continued to persist. Under such circumstances the team contacted Mr. Andrew Tridgell who had lost one of his planes due to a similar problem . With his generous help the team then accessed Pixhawk’s Log files to determine the exact cause of the problem. It was found that the use of digital servos led to frequent spikes in the voltage which when encountered, sent the autopilot into a complete reboot process.  Subsequently after  thorough root cause analysis of problem with support from Mr. Andrew Tridgell we learnt that voltage spikes can be negated by replacing some of the digital servos with analog or by using Zener Diodes. Ultimately the Pixhawk system turned out to be a huge success for the entire team. I think our journey of learning with Pixhawk has been commendable and exquisite ! 

This year team Edhitha is again participating at the SAUS 2015 and planning to incorporate the pixhawk autopilot again into our system. Now having mastered Auto-Nav on  Pixhawk we would like to explore features of Pixhawk in domains of Gimbal control, Antennae tracking, package drop, auto - camera triggering and most importantly we are working on interfacing the Image processing platform (single board computer) and pixhawk. 

3689637677?profile=original

3689637624?profile=original

 

3689637484?profile=original

Read more…