Hacker News Reader

Comment by kcorbitt | Hacker News Reader

I'm no control theory expert, but I have a hard time imagining how you could control a quadcopter in 3+ degrees of freedom with only a single input (the speed of your single remaining motor) no matter how fancy your algorithms get.

However, as unwind mentions, the researchers in the article claim it's possible.

View on HN

The essence of the algorithm is that you spin in the horizontal plane. If you have enough power your single remaining motor can create a vertical airstream that will keep the altitude constant. The limit on spinning is the structural integrity of the system with all that angular momentum, the limit on flying is the frequency response of your control system.

A number of people have built and demonstrated so called 'half prop' helicopters based on the same principle. If you have ever seen an maple seed fall you can understand the principle, search for 'maple copter' on Youtube for some videos of them in action.

Example: http://www.livescience.com/12725-maple-copter-robotic-craft-...

With one motor, if the vehicle is spinning, you could perhaps have a burst of increased thrust at a particular point in the spin and use centripetal force to move forward.

The article mentions that 'full control' becomes increasingly unlikely as you lose rotors, but partial control is still possible. You can influence how fast the quadrocoptor falls, reducing impact velocity or steering it to crash in a more favorable location.

Reply on news.ycombinator.com