Detailed screen field customisation

The ‘Voice Fields’ node is used to voice-enable an RF screen. It is a sub node of the ‘Logical Step / Screen’ node.

Individual fields within a screen can be added and sequenced to define interactions required throughout the voice flow (e.g. conditions, prompts, inputs).

Voice Fields Configuration

Defining the field sequence

All the fields in a screen relevant to the voice flow should be added to the table and sequenced with a ‘Seq No’. The PreBilt framework evaluates each field in ascending sequence to determine an action, whether that is outputting a spoken prompt and moving on or requesting a user input.

The sequence number should be unique for each field in the table.

Field Seq No Definition

Assigning a voice input condition

Each field added in the ‘Voice Fields’ node must be assigned a Voice Input Condition. This controls when the field is relevant during the screen voice flow.

Assigning a Voice Input Condition

Refer to Voice Input Conditions for a list of available voice input conditions.

Example

If a field is display-only but it is always necessary for the system to read out loud its value, you can use Voice Input Condition = ALWAYS or FOCUS and Voice Input Method = SpeakPrompt. Even if the focus is not set on the field, the system will speak the defined prompt when it is next in sequence.

The screenshot below shows a proposed customising for the WO details screen where the system will read the WO number, the number of pick HUs and number of tasks. This can be useful to provide a summary of the pick to the operator when this screen is called.

For instance, “Warehouse order 2, 2, 4, 0, Pick HUs, 1, Tasks, 3”.

Using Voice Input Condition ALWAYS

Outputting a voice prompt

Any voice prompt text can be assigned to a screen field, which will be spoken when the field is next in the sequence. This can be used to provide instructions to the warehouse operator, or to request an input or confirmation.

Voice Prompt Text

A voice prompt text can also contain any voice field value if it is available in the current screen’s structure. The complete field name must be contained within curly brackets { }.

Custom Data Fields

If a new custom data field is required within a voice workflow, an ABAP developer must:

• Add the new field to the required screen structure.

• Create a copy of the function module used in the corresponding logical transaction step/function code in the standard RFUI customizing.

• Add the new logic to populate the custom field in the function module.

• Assign the new function module against the corresponding personalization profile/logical transaction step/function code in the standard RFUI customizing.

The screen dynpros do not need to be modified to add the new field. As long as it is appended to the RF structure, it will be available for use in voice prompts.

Using functions in voice prompts

Special PreBilt function modules can be used to dynamically convert or format a field value in a voice prompt. These can be easily called by adding the required function name* before the field name, followed by a colon.

Note

Each function expects a certain field parameter. A valid field name must be used against a function to prevent unexpected errors or behaviours.

Example

The system must only only say the last 5 digits of a HU (separated by commas). Use the last5 function to achieve this:

{last5:/SCWM/S_RF_ORDIM_CONFIRM-PICKHU}

This converts the PICKHU value from 00000000000100410308 to 1, 0, 3, 0, 8.

The following standard functions are available in PreBilt:

Function	Use	Expected Parameter
alpha	Strips leading zeros from a numeric text.	Numeric
bin	Formats a storage bin name based on the bin name pattern configured in section 5.2.3.	Storage bin
decimal	Outputs a decimal value in user format, i.e. removes zeros to the right.	Numeric
last5	Returns the last 5 digits of a number, separated by commas.	Numeric
mat	Returns the product description using the Voice Language Key configured in section 5.2.1. If a product has no description maintained in the Voice Language, the default product description is returned instead.	Product number (MATNR)
uom	Returns the unit of measure description.	Unit of measure

New functions can be developed in customer implementation projects to cover further specific requirements.

Assigning a voice input method

Each field added in the ‘Voice Fields’ node must be assigned a Voice Input Method. This enables the system to react correctly based on the expected input type.

Assigning a Voice Input Method

Voice input methods starting with Get* only expect a certain data type, such as alpha, numeric or float.

ReadyPrompt can be used when the expected input for a field prompt is just “ready”.

SpeakPrompt only speaks the required prompt and moves on (i.e. no input is required).

Disabling auto back

By default, the “back” voice command is always available (so long as BACK or BACKF is available in the RF step), which lets voice operator invoke this command at any point to return to the previous navigation screen. However, in specific situations, this may not be necessary.

The ‘No AutoBck’ checkbox can be flagged to disable the “back” command for any field in the screen as required. This will also remove it from the “talkman help” options.

Disabling Auto Back

Submitting a command after input

A command such as INT_ENTER or INT_FCODE can be submitted after any voice field input.

Submit Voice Command

With INT_FCODE, any function code previously configured in the standard RF framework can be submitted after the field input, by adding the function code as a parameter in the ‘Commands’ sub node.

Example

A new function code ZHUCR has been configured in the Pick HU screen, which calls the standard HUCR + NEXT at once. This needs to be called right after inputting the Pick HU to initiate picking.

• Assign Submit Voice Command = INT_FCODE to the Pick HU voice field.

• In the ‘Commands’ sub node, create an entry for Voice Command = INT_FCODE and Type = ‘FC’, with Submit = ‘X’ and Voice Command Parameter = ‘ZHUCR’.

Submitting an FCODE voice command

Assigning voice commands

Specific voice commands can be made available for any field in a screen, whether they are used to provide additional information to the user or to submit a function code.

A voice command with Type = FC or OC can be used to call a function code. The FCODE must be assigned in the ‘Voice Command Parameter’ field.

Calling an FCODE with a Voice Command

• Voice Command Type = FC will submit the function code immediately.

• Voice Command Type = OC will ask for a confirmation before submitting the function code. This is useful to prevent the talkman from accidentally detecting the wrong voice command.

Example

If you assign Type = OC to a voice command for the word “save”, the default confirmation prompt in this case will be: “You said save, correct?” (yes/no)

• Voice Command Type = VO can be used to provide information to the user (e.g. query a field). The voice prompt must be added to the ‘Command Parameter Text’ field. Functions can also be used here to dynamically convert or format field values.

• 

  Providing Information to User with Voice Commands

Configuring the Talkman Help voice prompt (INT_HELP)

VoiceDirect’s “talkman help” voice command can be invoked by the user at any point during a voice flow. This provides additional instructions and guidance in case the operator is unsure what to do next, or does not know which voice commands are available during the voice flow.

Under the ‘Commands’ sub node, the Voice Command = INT_HELP with Type = HE can be used for any voice field to define the exact prompt when “talkman help” is invoked.

INT_HELP Voice Command

“Talkman help” will also say any voice commands that are assigned in the voice field with Type = VO or FC.

Example

The configuration entries in the screenshot above will result in the following “talkman help” prompt:

“Go to the source bin and read the label with check digits, or say back or pick HU or bin”

Note voice commands “pick HU” (VCMD0000002) and “bin” (VCMD0000004) are also added to the prompt, as they are configured in the screen field’s commands with Type = VO.

Using extra arguments

Additional arguments can be used in the voice flow configuration to cover special scenarios, such as outputting validation prompts after a user input or restricting the min/max length of a field input.

Extra Arguments

Arguments CNF_VAL_PRMT_POS, CNF_VAL_PRMT_PRE and CONF_VALUE are required in a voice flow where it is expected for the system to validate a spoken user input, before passing the input to the RF framework.

Example

After the user says “3” to confirm a picked quantity, before passing the input to the screen field, the system is meant to validate the input by confirming “You said, 3, correct?” and expect a “Yes” or “No” answer.

This can be achieved by defining the following extra arguments:

• CNF_VAL_PRMT_PRE = “You said”

• CNF_VAL_PRMT_POS = “correct?”

• CONF_VALUE = ‘X’ (expects a Yes/No confirmation)

Arguments MAX_LENGTH and MIN_LENGTH can be used to ensure VoiceDirect is only expecting a certain length range for a particular input field. This helps in minimising input errors.