Hobbyist Friendly, Autonomous Drone...

3689620455?profile=original

Greetings fellow drone enthusiasts…

I wanted to share with this community a project I've been working on for nearly 2 years now...

 

Drones are offering unprecedented aerial perspectives to amateurs, but their high price and often steep learning curve make them inaccessible to may hobbyist that could be aided by them.  

This project is intended to develop a drone to compliment people's current interests and hobbies, while not replacing them.  That requires a drone which is reliable, easy to use and can be thrown in an equipment bag, the back seat of a car or quickly tethered to a backpack without a second thought.

 

Inspiration Strikes...

About 5 years ago my brother and I moved from Massachusetts to Arizona for an aerospace job in The Valley.  We had always enjoyed New England's beautiful forests and top notch climbing destinations, so the idea of being surrounded by craggy mountains and winding canyons was too good to pass up.  One particular destination that caught our attention was Granite Mountain in Prescott AZ, a 500 ft cliff of Proterozoic granite surround by low laying shrubs, conifers and junipers.  3689620426?profile=original

Excited to scale this crag, we packed our climbing gear and set out on the 2 and a half hour drive from Phoenix to Prescott.  Before long we were hiking along the well marked trails of Prescott National Forest.

As many climbers know, when approaching a cliff there comes a time when you must leave the well marked and maintained hiking trails and pick your way along the often confusing and poorly marked climbing access trails that lead to the base of your climb.3689620470?profile=original  Granite Mt was no difference.  After a few hours of hiking we knew we should be close to the access trail, but in the high desert, surround by low shrubs, towering boulders, washes and game trails we realized that identifying and following the trail would be no small feat.  Time and time again, we started up what we thought was the correct path only to back-track when the trail went cool.  Eventually, the afternoon monsoons began gathering in the distance and it became obvious that this trip would end like many others, a lengthy, frustrating hike with 40+ pounds of gear that would not be used.3689620388?profile=original  

There would be 2 more attempts at Granite Mountain before we found the right path and completed our climb, but once at the top there was a realization.... Looking back down over the approach from 500 ft above, the access trail was obvious.  The bird's eye view clearly showed a long winding path unobstructed by boulders or spiny shrubs leading right from the hiking trail to the base of the climb.  We knew what we needed the next time our climbing mission was threatened by a confusion approach; a compact, autonomous drone capable of providing aerial snap shots of our surroundings.

Leveraging our aerospace backgrounds, our skills with embedded electronics and 3D modeling/printing we started on a drone which would be free from what we saw as their current limitations and created a drone which would be rugged, inexpensive, compact and intuitively work without any prior experience in R/C hobbies. 

Configuration...3689620514?profile=original

From the beginning we knew that the traditional quad or multi-copters would be challenging to pack due to the many arms and propellers protruding from their bodies. Also, each arm requires a separate motor and speed controller adding to the total weight, expense and complexity of the drone. 

To overcome these issues, the configuration of our drone is based on a co-axial, fixed pitch helicopter.3689620483?profile=original This configuration allows for a tighter profile and requires only two motors and speed controllers which can be mounted safely inside the main body.  To further reduce the overall size and allow the optimum positioning of sensors we wanted to utilize the space not just below the plane of the rotors, but between and above them as well.  This was realized by mounting the rotors on a fixed, hollow, carbon fiber shaft (which we refer to as the “spine”).  The spine runs the length of the drone and not only allows routing of electrical cables past the planes of the spinning rotors but gives the drone wonderful stiffness and strength.3689620444?profile=original

Propeller blades offer another challenge to packing a drone. They tend to be fragile when bent and will catch and snag on other equipment.  To overcome this, our drone utilizes torsion springs in the blade’s flapping hinges to keep them tightly folded along the profile of the drone when not spinning.  In flight, however, the centripetal force of the spinning rotors overcomes the springs and allows the blades to extend outward.  Testing has shown that these torsion springs have negligible effect on the dynamics of the blades even at low RPM.

Mission Overview...3689620497?profile=original

The Drone is designed to be a fully autonomous and capable of flying a variety of missions which are pre-programmed by the user, using a mobile app and uploaded via Bluetooth.  Once the mission is selected/defined and uploaded, the drone is hand launched to commence its mission: flying a circuit, capture images or sending a text message from a better vantage point than the user on the ground.

To increase the packageability of the drone, there are no landing features which protrude from its body.3689620533?profile=original  Instead, power is cut to the motors and the torsion springs mounted in the blade’s flapping hinges fold the blades along body allowing it to dive towards the ground.  While falling, the drone tracks its altitude and descend rate and when certain conditions are met, power is restored to the rotors extending the blades and braking the fall.  Finally, just prior to impact with the ground, power is again cut, the rotor blades fold into their retracted position and the drone makes a soft impact with the ground.  Mission details and images are then downloaded to the mobile app.


Current Status of the Project...

The first cut at this drone was intended to prove out several fundamental concepts such as the electrical routing, gear ratios, 3D printed design, flybarless stability and the sensor’s resistance to vibrations.  However, because it still lacked several key features it was difficult to call a prototype and instead was referred to as the “Flying Test Stand”.  Eventually more features such as GPS/barometer, ultrasonic range finder, torsion springs, aeroshells, etc. where added.  This allowed us to begin outdoor testing to investigate ascent rates, GPS tracking and hand launching all of which were successfully demonstrated.  Shortly, testing will begin on giving the drone a greater level of autonomy and stringing together multiple mission events while also increasing the durability and robustness of the drone. 

Hope to update everyone with videos the more progress soon!!!

3689620569?profile=original

3689620616?profile=original

3689620549?profile=original

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • Clever design. Awesome form factor. I look forward to seeing this design develop further.

  • I published this post a few times with minor changes not understanding it would delete comments.  Think a few were lost, my apologies.

  • Oh, also given that you're carrying a camera, you could ditch the IMU stuff entirely and run with a "visual-IMU" system based on the University-of-Zurich ideas published here recently

  • Interesting concept.

    I'd ditch the swashplate control and attempt a weight-shift mechanism myself. 

  • Moderator

    I do believe I left a comment, I do hope no others were removed from this post. We do not permit that here. 

  • Do you have a mailing list for updates?  I'd like to keep in the loop for your flying minion.  Any idea as to MSRP yet?

  • Hey Brian, yes it is all basically custom.  We're currently using a collection of development/breakout boards offered by places like Sparkfun.com or Adafruit.com.  We hand-etched some boards to help us cut down on the number of wires. The coding is in C and while we barrowed as much as we could from open source libraries (for writing to the SD card, managing serial communication, etc) the software is largely our design.

    We are currently moving away from purchasing breakout/development boards and looking to have our own board layout manufactured by an outside house.  This would let us make the electronics smaller and hopefully cheaper.

  • The original concept was just something to pop-up and snap some photos before falling back to the user, so flight time wasn't really a concern.  After we opted to include a GPS to limit its wondering during the ascent, having it fly simple circuits was not much extra work.  My point is we've never really seen it as a long endurance, loitering surveillance drone so not much time was spent testing flight time.

    That being said, we currently use 2, 7.4v brushless motors with 18amp ESC and a 1300 mAh battery.  If I remember correctly we pull about 15 to 18 amps (split between the ESCs) at 75%-80% throttle (which is our current setting for the ascent).  So if I do the math right that's about five mins, but hard to comment on if that's right or not.  The drone just doesn't use that much motor. It's on for about a 15 sec to 20 sec burst, can shut off and ride that momentum for additional altitude and is generally off during the majority of the decent.  If it orbits to capture some pics away from the user it's at reduced power and that too is only for 30 sec or so as it flies over and back. 

    So perhaps a more relevant unit of measure is number of missions between charge, which is about 6 to 8 right now.

  • I can see now, hard to tell at size and resolution shown.

    Great concept, the larger swept rotor area should make it more efficient than a quad or other multi and the compact carrying arrangement should make it a lot more practical for people to carry around with them.

    Can't wait to see them for sale.

    Best Regards,

    Gary

  • Very impressive!  What flight controller / software are you using?  Is it all custom?

This reply was deleted.