This section describes the AI Controller specifications that are unique to the AI Controllers.
NY-series Artificial Intelligence Controller
last update: April 1, 2021
This section describes the AI Controller specifications that are unique to the AI Controllers.
Item | Description | |
---|---|---|
AI Functions |
Time Series
Database Function |
The Time Series Database Function collects values of user-specified variables and
calculation results of the Feature Value/ Machine Learning Function into the storage for each sampling interval. This function allows you to collect data such as variable values to the storage without program. The collected data can be checked on the AI Viewer. In addition, the data can be transferred to a web server by the WebAPI Connection Function. |
Feature Value/
Machine Learning Function |
The Feature Value/Machine Learning Function determines whether equipment events
occur from the collected data and AI machine learning model. It consists of the Feature Extraction Function and the Machine Learning Function. The Feature Extraction Function calculates feature values from data. The Machine Learning Function on the other hand is designed to determines whether equipment events occur from feature values and AI machine learning model. |
|
WebAPI
Connection Function |
The WebAPI Connection Function transmits data (CSV files) that is collected by the
Time Series Database Function in the AI Controller to a web server periodically. This function can be used to transfer files that is collected by the Time Series Database Function to a web server and to save and analyze data. Considering that data will be transmitted via Internet, data is encrypted in the transmission path to the web server. |
TimeSeries | Item | Specifications |
---|---|---|
TimeSeries
common |
Method | Time-series database
(It uses a circular queue where the oldest data is deleted if it reaches the maximum number of data.) |
Number of TimeSeries | 4 | |
Sampling start/stop
method |
It can be executed in any of the following methods:
• AI Operator • System-defined variables • Instructions |
|
Export start/stop method | It can be executed in any of the following methods:
• AI Operator • System-defined variables • Instructions |
|
Number of used variables
with a Retain attribute |
2 *1 | |
Variable data
(RAW_DATA) |
Size of the TimeSeries | NX-series: 900 MB
NY-series: 41 GB |
Number of variables
[variables/record] *2 |
1024 | |
Category of variable | Global variable | |
Variable type | The following variables can be specified:
• Basic data types • Specifying array elements: • Specifying members of a structure or a union |
|
Analysis data
ANL_DATA) |
Size of the TimeSeries
database |
NX-series: 1 GB
NY-series: 30 GB |
Number of variables
[variables/record] *2 |
2048
(including variable data, frame variables, subframe variables, and label variables) |
|
Category of variable | Global variable | |
Variable type | The following variables can be specified:
• Basic data types • Specifying array elements: • Specifying members of a structure or a union |
Item | Specifications | |||
---|---|---|---|---|
Number of equipment events | 128 max. | |||
In each
equipment event |
Frame variables |
Number of variables that
can be registered |
1 | |
Supported data type | SINT, INT, DINT, LINT, USINT, UINT, UDINT, ULINT | |||
Feature extraction
out- put frame variables |
Number of variables that
can be registered |
1 | ||
Supported data type | Same types as the frame variables | |||
Feature value |
Number of variables that
can be registered |
16 max. | ||
Per
feature value |
Variable
data |
Number of variables that
can be registered |
1 | |
Supported data type | LREAL, BOOL | |||
Subfram
variables |
Number of variables that
can be registered |
1 * | ||
Supported data type | BOOL | |||
Machine learning
output frame variables |
Number of variables that
can be registered |
1 | ||
Supported data type | Same types as the frame variables | |||
Number of classifications for equipment event
monitoring |
3 (Normal, Alarm Level 1, Alarm Level 2) | |||
Equipment event detection algorithm | isolation forest |
* Up to six subframe variables can be registered to an equipment event.
Item | Specifications | ||
---|---|---|---|
Function
specifications |
Execution trigger | File transfer can be executed in the period specified by a
user on the AI Operator. |
|
Destination specification | Specify a URL of the server to which files are transferred.
Specify a URL starting with http:// or https://.When you specify https://, SSL/TLS communications are established. |
||
File deletion after transfer | Once the file has been transferred to the Web server
successfully, the WebAPI Connection Function deletes the file in the AI Controller. |
||
Connection check function | To check the connection with the Web server, a file transfer
can be triggered by the AI Operator at a given timing. Refer to NX/NY-series Artificial Intelligence Machine Automation Controller User’s Manual (Cat. No. W594) for details. |
||
Transfer specifications | If there is more than one file to transfer at the transfer timing,
all the files are transferred. If the last file transfer is in progress and the next file transfer timing arrives, the execution of the last file transfer continues. |
||
Transfer protocol | The file transfer executes in accordance with the RFC1867
procedures. The multipart/ form-data in the form is used. The tag name “datafile” is used for the transfer. |
||
Communication
specifications |
HTTP/HTTPS
client |
Supported versions | 1.1 |
DNS | Supported | ||
Proxy | Supported | ||
Basic authentication | Supported (Basic authentication for the proxy server and Web
server connection is supported.) |
||
Number of files that
can be transferred simultaneously |
3 | ||
Security
(when https:// is specified for the address) |
TLS version | 1.0, 1.1, 1.2 | |
Server certificates | Import the certificates to the AI Controller with the AI Operator.
Up to 32 certificates can be set. |
||
Revocation check
for the certificates |
Revocation is checked by OSCP stapling. |
The specifications of the AI Controller other than the specifications described in Specifications Unique to the AI Controllers are in common with those of standard CPU Units or IPC Machine Controllers. Refer to the corresponding specifications for each AI Controller model according to the table below.
AI Controller model | Corresponding standard model |
---|---|
NY532-Z500 | NY532-1500 |
NY532-Z400 | NY532-1400 |
NY532-Z300 | NY532-1300 |
NY512-Z500 | NY512-1500 |
NY512-Z400 | NY512-1400 |
NY512-Z300 | NY512-1300 |
Refer to the following manual for the performance specifications.
• NY-series AI Controller: NY-series Industrial Panel PC / Industrial Box PC Software User’s Manual (Cat. No. W558)
Specifications of non-AI functions are same as those for the standard CPU Units or for the IPC Machine Controllers without the AI functions. Refer to the following manual.
• NY-series AI Controller: NY-series Industrial Panel PC / Industrial Box PC Software User’s Manual (Cat. No. W558)
last update: April 1, 2021