At joint level, we now typically have Maxon or Minimotor DC motors. The position is actively controlled on the basis of quadrature signals generated by optoelectronic sensors mounted on motor axis. 
 
At joint level, control algorithms consist in simple, traditional PID loops. Optionally, a priori components relating voltage or current to speed and acceleration could be added in order to eliminate tracking errors. But for our applications we can easily limit errors to the order of 1 per thousand without these additional measures. 
 
More critical factors, which we rely on, include an adequate motion law at coordination level, the management of absolute values for location, and the appropriate management of maximum control intensities. 
 
Position control is performed by specialized controllers. While we typically use Galil controllers, we have also designed solutions with Fiveco, iXs, and K-Team. The use of a PLC (in our case typically a Beckhoff system) could also be considered now (traditionnally this could be found only for much higher power, such as for 1kW and more) 
 
Previously we used stepper motors, which allow for open-loop control, i. e. spare the need for sensors and fast computing resources. Stepper motors perform with very contrasting success: depending on design and application, they can be "perfect" (no static problem nor tracking error) or cannot be used at all (important losses of steps). For moving structures however, their mass five times larger is often found a drawback. Power consumption may also comparatively be a disadvantage. 
 
 
>> next - Motors and drives level  
 
 
<< more general - Motion control