You are on page 1of 13

Module 3

Control and Display Technologies

Direct Voice Input (DVI)


Mohamed Sameer T K Dept. of Aeronautical JCET

Direct Voice Input (DVI)


Direct voice input (DVI) control is a system which

enables the pilot to enter data and control the operation of the aircrafts avionic systems by means of speech.
The spoken commands and data are recognized by a

speech recognition system which compares the spoken utterances with the stored speech templates of the system vocabulary.
The

recognized commands, or data, are then transmitted to the aircraft sub-systems by means of 2 the interconnecting data bus.

DVI
As examples:
(a) To change a communication channel frequency,

the pilot says radio (followed by) select frequency three four five decimal six.
(b) To enter navigation data, the pilot says

navigation (followed by) enter waypoint latitude fifty one degrees thirty one minutes eleven seconds North. Longitude zero degrees forty five minutes seventeen seconds West.
3

DVI
Feedback that the DVI system has recognized the

pilots command correctly is provided visually on the HUD and HMD (if installed), and aurally by means of a speech synthesizer system.
The pilot then confirms the correctly recognized

command by saying enter and the action is initiated.

DVI
The pilot can thus stay head up and does not have to

divert attention from the outside world in order to operate touch panels, switches, push buttons, keyboards etc.
DVI can thus reduce the pilots work load in high work

load situations.
The pilots are trained to speak clearly and concisely

in a strongly structured way when giving commands and information over the communication channels to fellow crew members, other aircraft and ground 5 control.

Characteristics and Requirements for an Airborne DVI


Fully connected speech: The speech recognition

system must be able to recognize normal fully connected speech with no pauses required between words. (Systems which require a pause between each word are known as isolated word recognizers).
Must be able to operate in the cockpit noise

environment. The background noise level can be very high in a fast jet combat aircraft.
Vocabulary size. The required vocabulary is around

200 to 300 words.


6

Characteristics and Requirements for an Airborne DVI


Speech template duration: The maximum speech

template duration is around 5 seconds. Vocabulary duration: The maximum duration of the total vocabulary is around 160 seconds. Syntax nodes. The maximum number of syntax nodes required is about 300.
An example of a typical syntax tree is shown below

Characteristics and Requirements for an Airborne DVI


Duration of utterance: There must be no restrictions

on the maximum duration of an input utterance.


Recognition response time: This must be in real

time.

DVI
The basic principles are to extract the key speech

features of the spoken utterance and then to match these features with the stored vocabulary templates.
Sophisticated algorithms are used to select the best

match and to output the recognized words.


Very extensive research and development has been

carried out and is a continuing activity worldwide to produce speech recognition systems which are speaker independent, that is, they will recognize words spoken clearly by any speaker.
9

DVI
Recognition accuracies of at least 96% is required in

the cockpit environment. In the case of numerical data, numbers which are outside the likely range for that quantity, for example, radio frequencies or latitude/longitude co-ordinates can be rejected.

10

Display Integration with Audio/Tactile Inputs

11

Display Integration with Audio/Tactile Inputs


The integration and management of all the display

surfaces by audio/tactile inputs enables a very significant reduction in the pilots workload to be achieved in the new generation of single seat fighter/strike aircraft.
The effectiveness of the system in reducing pilot

workload when combined with the carefree maneuvering resulting from the FBW flight control system and the automated engine control system is referred to as Voice, Throttle, Stick control.
12

DVI
Problem with Voice Recognition
Voice control is not suitable for time critical system.

The words in the vocabulary are limited.


Generating templates are time consuming. Microphone have the same electrical characteristics

as the flight microphone Speaker independent Speech Recognition requires large amount of memory ,and large signal processing

13

You might also like