Page 3
ASUS, Inc...
2 2021 ASUS, Inc... All rights reserved...
2... 4 Verification ... 16
2... 4... 1 Configure task and parameter ... 16
2... 4... 2 Select the data source ... 17
2... 4... 3 Select the data for verification ... 17
2... 4... 4 Adjust Threshold ... 18
2... 4... 5 Confusion Matrix ... 20
2... 5 Testing ... 22
2... 5... 1 Select model ... 22
2... 5... 2 Result Pass/Fail ... 23
Chapter 3: Upgrading AISDetector ... 24
3... 1 Complimentary software upgrade policy ... 24
3... 2 How to upgrade AISDetector ... 24
3... 3 Cost to upgrade AI Software ... 24
Chapter 4: Support for your AI software ... 25
4... 1 Before you call customer support ... 25
4... 1... 1 Troubleshooting ... 25
4... 1... 2 Call for support ... 25
Appendix ... 26
Safety Information ... 26
Regulatory notice ... 26
Contacting ASUS ... 26
Page 4
ASUS, Inc...
3 2021 ASUS, Inc... All rights reserved...
About this manual
This manual provides an overview of the key features of this software and gives step-
by-step instructions for making full use of them...
How this manual is organized
Chapter 1 Getting to know AISDetector
This chapter gives an overview of the key features and functions of the
software...
Chapter 2 Using AISDetector
This chapter provides information on how to use the software...
Chapter 3 Upgrading AISDetector
This chapter provides information on how to upgrade the software,
pre-trained models, and other necessary functions...
Chapter 4 Support for AISDetector
This chapter provides information for troubleshooting and contacting
ASUS for support...
Appendix This section includes the notices and safety statements for this
software...
Page 5
ASUS, Inc...
4 2021 ASUS, Inc... All rights reserved...
Conventions used in this manual Throughout this manual, blocks of text as shown below are used to emphasize
important information...
IMPORTANT! This message contains vital information that must be followed to
complete a task...
NOTE: This message contains additional information and tips that can help to
complete a task...
WARNING! This message contains important information that must be
followed to keep you safe while performing certain tasks and prevent damage
to AISDetector data and functions...
Typography Bold Indicates a menu or an item to select
Italics Indicates a section in this manual with additional information
Package contents Inside the package is a USB dongle, which contains the activation key for AISDetector...
Page 6
ASUS, Inc...
5 2021 ASUS, Inc... All rights reserved...
Chapter 1: Getting to know AISDetector
1... 1 Introduction
AISDetector is an Artificial Neural Network (ANN) based application software that
together with a unique ASUS Artificial Intelligence (AI) user interface is designed to
detect anomalies in sound and vibration... The ability to detect these anomalies on the
factory production floors allows manufacturing challenges to be addressed promptly
for time and cost savings...
1... 2 Software
AISDetector is an AI-based end-to-end signal anomaly detection software for
collecting signals, training models using data sets, verifying and testing models, and
presenting results visually... AISDetector detects anomalies to help determine whether
products or equipment are running in a normal state...
With a friendly visual user interface (UI), AISDetector is extremely easy-to-use and
comes with a series of embedded tutorials that can be accessed at any time...
Simply label training signal data as OK or NG for various options of hardware type for
AI training without the need for data pre-processing... By following the uniquely pre-
defined ASUS AI software workflow, you can quickly train, verify, and test the target
to get results in a matter of minutes and easily become a second-to-none signal
analyst... With the data waveform displayed in the UI, you can narrow down the root
cause of any anomaly and conduct further analysis for process optimization...
1... 3 System Requirements
Your computer must meet the minimum system requirements below to run the
artificial intelligence model training and inference functions in this software...
CPU: Intel Core i3 or higher
DRAM: 8GB or more
USB: 2 free USB ports (For the license dongle or USB microphone)
Storage device: SSD (recommended)
Storage storage: 800 MB
Operating System: Windows 10 (64 bit)
Page 7
ASUS, Inc...
6 2021 ASUS, Inc... All rights reserved...
1... 4 Features
New target/project set up with just a few clicks
Using fast AI modeling technology, users can create a new AI model within 1
min, and this makes production-line changeover easy...
Designed to work seamlessly with sensors for real-time data collection and
testing
Dataset source flexibility - accepting either sound from a microphone or
vibration from an accelerometer
Variable frequency source allowable and anomaly detection with time-
frequency techniques
No AI accelerator needed: AISDetector only requires computing power from a
typical CPU... This means that an AI-based manufacturing system can be
adopted without significant investment...
1... 4... 1 Training
The training module includes Project, Train, Verification, and Testing for you to create
your own AI project...
1... 4... 2 Inference
In this section, you can create, configure, or import a project for various detection
targets...
Page 8
ASUS, Inc...
7 2021 ASUS, Inc... All rights reserved...
Chapter 2: Using AISDetector
2... 1 Download the software
Go to the ASUS IoT website and find the AISDetector product download page...
2... 1... 1 Install the software
Run AISDetector... msi once the image file is downloaded...
2... 1... 2 Activate the software
When prompted for the activation key, plug the ASUS IoT USB dongle into the
computer that you have installed the AISDetector...
2... 2 Project
This section is to help you create, configure, or import a project for each different
detection target... You can select the data source angle and define the hardware
category for a new AI project to be used for subsequent steps...
Add new project information
Define data saving path
Create or import a project
Configure detailed setting
Page 9
ASUS, Inc...
8 2021 ASUS, Inc... All rights reserved...
2... 2... 1 Add new project information
Allows you to name the project, add a description (optional), and select a
hardware type in the red box below...
2... 2... 2 Define data saving path
Allows you to define and manage test data location...
Click the sign at the top right corner to access project settings, select the Path
tab, set the path, and click OK...
Page 10
ASUS, Inc...
9 2021 ASUS, Inc... All rights reserved...
2... 2... 3 Create or import a project
Create a new project or import a previously defined or an existing project...
When all information is successfully selected in the Create Project box, click ADD
at the top right corner to create the new project...
To import an existing project, click Import at the top left corner...
Page 11
ASUS, Inc...
10 2021 ASUS, Inc... All rights reserved...
2... 2... 4 Configure detailed setting
Once a project is created or imported, the corresponding project information will
be shown in the Project List that is boxed in blue...
When you select a project, the project details are shown boxed in green... The pre-
trained model will be shown boxed in yellow as a task count...
Project summary will be shown in the area boxed in orange once a project is
activated...
Page 12
ASUS, Inc...
11 2021 ASUS, Inc... All rights reserved...
2... 3 Train
In this section, you are walked through the training process from creating a new
training task to finishing the training...
2... 3... 1 Create a new training task
Allows you to create a new task in Task List by entering a new task name and
description (optional)...
Click the Create new task icon in Task List... Enter a new task name and
description (optional), and then click OK...
Create a new training task
Modifiy or delete task
Select training
data source
Set training parameters
Start training
Complete training and check result
Page 13
ASUS, Inc...
12 2021 ASUS, Inc... All rights reserved...
2... 3... 2 Modify and delete task
Allows you to modify and delete task information
Click the Pen icon in Task List to make modifications... When done, click OK...
To delete a training task, select the task that you want to delete, click the Recycle
Bin icon in Task List...
Once the setup is complete, click Training Data, which is shown boxed in green
below, to select the training data source...
Page 14
ASUS, Inc...
13 2021 ASUS, Inc... All rights reserved...
2... 3... 3 Select training data source
You can select training data from the following three (3) data sources...
1... Use on-site accelerometer...
2... Load or select single or multiple recording files
3... Import an entire folder of sound recordings...
For all sound recording using a microphone, you can define the recording angle...
Even after the setup is complete, you can still change the training parameters
prior to training if needed...
Click Start to initiate the training... During the training process, the loss curves on
the X, Y, and Z axes, as well as the elapsed and remaining time, are displayed in
the main pane...
Page 15
ASUS, Inc...
14 2021 ASUS, Inc... All rights reserved...
2... 3... 4 Set training parameters
Click Setting to modify parameters...
Click the Model tab in the Setting dialog box...
Update the training Epochs or Loops, and then click OK to save...
2... 3... 5 Start Training
Once the setup is complete with parameter settings, click START to start the
training...
During training, the loss curves for the X, Y, and Z axes, as well as the training
time and progress will be shown for reference...
Page 16
ASUS, Inc...
15 2021 ASUS, Inc... All rights reserved...
2... 3... 6 Training complete
Once the progress percentage shows 100%, training is complete and ready for
verification and testing...
Page 17
ASUS, Inc...
16 2021 ASUS, Inc... All rights reserved...
2... 4 Verification
This section takes you through the evaluation and verification of the model...
2... 4... 1 Configure task and parameter
Select the tasks (boxed in red), angle (boxed in green), feature (boxed in blue),
and threshold (boxed in orange), and then click Start to start the evaluation...
Configure task and parameter
Select the data source
Select the data for
verification
Adjust Threshold
Confusion Matrix
Examples of Confusion
Matrix
Page 18
ASUS, Inc...
17 2021 ASUS, Inc... All rights reserved...
2... 4... 2 Select the data source
Data source options include accelerometer on-site, load single or multiple source
files, or import data from a folder to verify detection, loss, and overkill results...
2... 4... 3 Select data for verification
You have the option to enable or disable each dataset for evaluation...
Page 19
ASUS, Inc...
18 2021 ASUS, Inc... All rights reserved...
2... 4... 4 Adjust threshold
You can also select a specific angle (All, X, Y, Z) for verification and further
evaluation to verify detection, loss, overkill results... You can refer to steps A to D
below...
Step A:
Step B:
Page 20
ASUS, Inc...
19 2021 ASUS, Inc... All rights reserved...
Repeat steps A through C until you have completed the threshold adjustment...
Steps C and D:
Page 21
ASUS, Inc...
20 2021 ASUS, Inc... All rights reserved...
2... 4... 5 Confusion Matrix
Refer to the figure below for the definition and explanation of the Confusion
Matrix...
True Positives (TP): Predicted target event a Positive, and the actual
event is a positive...
True Negatives (TN): Predicted target event a Negative, and the actual
event is a negative...
False Positives (FP): Predicted target event a Positive, and the actual
event is a negative...
False Negatives (FN): Predicted target event a Negative, and the
actual event is a positive...
As mentioned above, we can use TP, TN, FP and FN values to calculate
accuracy, recall, loss, detection and overkill:
Accuracy = (TP+TN) / (TP+FP+FN+TN)
Accuracy is the ratio of total sum of all true events predicted
over total sum of all predicted events...
Precision = TP / (TP+FP)
Page 22
ASUS, Inc...
21 2021 ASUS, Inc... All rights reserved...
Precision is the ratio of true positive events predicted over
total sum of all predicted positive events...
Recall = TP / (TP+FN)
Recall is the ratio of true positive events predicted over
total sum of actual positive events...
Loss = FN / (TP+FN)
Loss is the ratio of FN events over total sum of actual
positive events...
Detection = TP / (TP+FN)
Detection is equal to Recall...
Overkill = FP / (FP+TN)
Overkill is the ratio of false positive events over sum of total
actual negative events...
Examples of Confusion Matrix
Let's say, there are a batch of examinees total of 100 pcs... The true event is that
92 are good and 8 are no-good with defect...
The examination event is as below:
1... 5pcs examined with defects, and they are defect parts (True Positive)
2... 90pcs examined good, and they are actually good parts (True Negative)
3... 3pcs examined with defects, and they are actually good parts (False
Positive)
4... 2pcs examined good, and they are defect parts (False Negative)"
Hence, the result as below
1... Accuracy = (TP+TN) / (TP+TN+FP+FN) = (5+90) / (5+90+3+2) = 95%,
meaning 95% of total examination events are accurate result...
2... Precision = TP / (TP+FP) = 5 / (5+3) = 62... 5%, meaning 62... 5% of total
examined defect events are actually defect events...
3... Recall or Detection = TP/ (TP+FN) = 5 / (5+2) = 71... 4%, meaning 71... 4% total
actual defect parts examined as defect events...
4... Loss = FN/ (TP+FN) = 2 / (5+2) = 28... 6%, meaning 28... 6% total defect parts
are not examined as defect events...
5... Overkill = FP/ (FP+TN) = 3 / (3+90) = 3... 2%, meaning 3... 2% of total good
parts are examined as defect events...
Page 23
ASUS, Inc...
22 2021 ASUS, Inc... All rights reserved...
2... 5 Testing
This section shows how to use a model to test and for further evaluation...
2... 5... 1 Select model
Before performing any tests, double check that a model is selected and there is
equipment input... Once confirmed, click Start...
Select model Check Result-
Pass/Fail
Page 24
ASUS, Inc...
23 2021 ASUS, Inc... All rights reserved...
2... 5... 2 Result Pass/Fail
Real-time result (Pass/Fail) is shown on the screen for the model selected... You
can also select Show Plot to show the waveform for further analysis...
Page 25
ASUS, Inc...
24 2021 ASUS, Inc... All rights reserved...
Chapter 3: Upgrading AISDetector
3... 1 Complimentary software upgrade policy
You are eligible for a complimentary upgrade under the following circumstances...
A new version of AISDetector is released with bug fixes
A new version of AISDetector is released with new features/functions/pre-
trained models within 365 days of your AISDetector purchase
Otherwise, you will need to pay for an upgrade...
3... 2 How to upgrade AISDetector
Go to the ASUS IoT website and find the AISDetector upgrade page...
Click Upgrade, and the website will check whether your AISDetector is eligible for an
upgrade... If it is eligible for a complimentary upgrade, the system will automatically
perform the upgrade...
Once the upgrade is complete, re-install and re-activate again by following the same
instructions for first-time installation and activation of your AISDetector software...
3... 3 Cost to upgrade AI Software
Check with your ASUS representative for more details...
Page 26
ASUS, Inc...
25 2021 ASUS, Inc... All rights reserved...
Chapter 4: Support for your AI software
4... 1 Before you call customer support
4... 1... 1 Troubleshooting
Double check your hardware and software settings to make sure that they are set to
run as designed...
4... 1... 2 Call for support
Find the phone number of the ASUS IoT support center that is nearest to or most
convenient for you...
A log file will be generated and saved in the system file manager folder
(AISDetector_version\etc\Log)...
Page 27
ASUS, Inc...
26 2021 ASUS, Inc... All rights reserved...
Appendix
Safety Information
Regulatory notice
Contacting ASUS
Taiwan
Talk to us
[email protected] com
Call Us | Official Support | ASUS Global
China
| ASUS
Germany
Telefonische Produktuntersttzung
+49 (0) 2102 5789 557
Telefon sterreich
+43 (0) 1360 2775 461
USA
Chat with us
1-888-678-3688
https://www... asus... com/us/support
en