How to Perform Tasks
Another important issue to consider regarding ARMs is the way that they can be controlled by the user to perform tasks. This includes the input devices, the menu structures, and different modes of control.
For industrial robots, there are several different input devices
- • Input coordinates via computer
- • Control panels
- • Several kinds of (solid, but heavy) joysticks
People with disabilities in their arms/fingers would in general not be able to control a robot with one of these devices.
For our application, we could consider a joystick that can provide many independent control signals, like those used for many games (e.g., Xbox® or PlayStation®). The higher the number of independent control signals, the more movements of the ARM can be combined (e.g., moving the end-effector in the 3-D space while controlling its orientation). However, joysticks that generate many control signals are more difficult to handle, especially for people who have problems with their arms/hands. For our target group, the following input devices are usually considered:
- • Touch screens;
- • Microlight switches; and
- • Personal computer (PC) control (e.g., the user controls the ARM through software running in the computer).
Other input devices may be considered, although they require special attention:
- • Eye-tracking control: A disadvantage of eye-tracking control is that an end user needs to look both at a screen and at the end-effector.
- • Voice control: This is generally not useful for giving nondiscrete commands. Switching a light “on” or “off” is relatively safe to implement with voice. For something that involves movement, safety is important. You could implement “up,” but how is this movement stopped when the robot does not “hear” the stop command correctly? This would mean you need to implement an extra safety feature. You could use voice to select commands, but have a (microlight) switch to actually initiate the movement. Once you release the switch, the ARM must stop.
- • Brain- or body-robot interface: This is still in an early stage of implementation. Currently, these input devices require a high degree of concentration by the user and are thus tiring.
Combinations of these controls can be effective for individual end users; therefore, ARMs should be able to support some or all of these input devices.