IRIS+ — not ready for prime time

Recently, we were commissioned by a local Twin Cities company to do a study on drones being marketed for their ability to fly in a programmed “automated” mode and, in particular, their ability to follow a person on the ground to video record him/her while engaged in various “extreme” sports (i.e. BMX biking, skiing, skateboarding, etc.). This article summarizes some of our findings.

iris+2

After a lengthy study, we selected the 3D Robotic’s IRIS+ quadcopter as a good, mid-range vehicle that was heavily marketed for its ability to fly on a preprogrammed course and also its ability to fly via its “follow me” technology, which allows the drone to follow a user on the ground who is holding a telemetry unit containing a GPS receiver. The telemetry unit broadcasts the user’s GPS position to the drone, allowing it to position itself in real time.

Sounds good, right? Well — not so fast. We very quickly learned that this technology is far from the consumer-grade experience that is being marketed. Here’s why:

  • Quadcopters are evolving rapidly, but they are still a bit of a science project. While being marketed as “flight ready” out of the box, the user needs to do some light assembly, source and attach a gimbal and camera, acquire a Tablet PC and find / load the necessary software (application and drivers). There is also an investment in time to learn the basics of manual flight and how the autopilot software works before anything can really be accomplished.
  • We were particularly disappointed in the camera gimbal mechanism. This device keeps the camera steady while the quadcopter is in flight.  This particular gimbal (made by Tarot) is essentially a circuit board mounted to the outside of the vehicle. The electronics and fragile cabling are completely exposed (no housing) where they can be easily damaged by a rough landing. It also required a hex wrench to attach/detach the GoPro camera, which needed to be done all the time to transfer video files and to charge the camera’s battery.
  • The software loaded onto the Android tablet (we used a Nexus 3) was the Tower application (also known as Droid Planner 3). At first glance, it appeared to be well constructed and fairly straightforward to use. However, we found out that it had VERY specific steps that needed to be followed in a very particular sequence in order to preprogram a flight plan that were not intuitive.   We learned that deviation from this pattern, while not obvious, can create disastrous results.
  • On a few occasions, we crashed the drone — not unexpected. However, we did learn that you are on your own for repairs. 3DR does not repair the product that it sells.       They do sell most of the parts for DIY repairs. However, the IRIS+ is clearly not designed to be repaired. We needed multiple tools just to get the covers off to expose the electronics and in one case we needed to unsolder wires to replace a broken plastic support arm. Definitely not user friendly even for experienced repair techs.
  • Lastly, we found lots of problems with their internal firmware in the flight control system. By far, the most annoying was a problem in which the IRIS+ did not detect a crash and continued to try to turn its props. This usually resulted in at least the destruction of the props and in one case caused the motor control printed circuit board to burn up (literally started on fire) when it overheated due to the excessive current being drawn from the stalled motors.

In summary, this mid-range quadcopter technology (and in particular the 3DR IRIS+ drone) is still very immature and requires a good deal of “tinkering” before it will do much of anything.   The control software is really in the same state — an open-source project that continues to evolve, but currently has many issues that can cause unpredictable results. And lastly, the design of the IRIS+ is marginal at best, with several firmware issues that need to be addressed before this can be called a stable, robust platform for the average consumer.

This technology will continue to advance to a consumer grade platform, but be advised it is not there yet.