Garbage In, Garbage Out
아무리 완벽한 알고리즘 모델을 구현하더라도, 좋은 품질의 데이터를 갖추지 못하면
머신러닝의 수행결과는 좋을 수 없습니다.
What We Do
에이치큐브는 알고리즘 모델 구현은 물론이고, 각종 센서를 통해 데이터를 수집하고 분석하여
시스템을 학습시키기 위한 최적의 데이터를 가공합니다.
Save Time, Cut Costs
최적의 데이터로 학습된 시스템을 통해 부품의 수명, 고장 등을 미리 판단하여
고객님의 시간 절약과 불필요한 지출을 피할 수 있게 도와드립니다.
생체 신호 모니터 및
혈압 예측 시스템
Customer : ETRI
Software : LabVIEW, Python
Hardware : NIBP(Non-Invasive Blood Pressure)
AD Instruments, Attachable Bio-signal Patch(ECG, BCG, Temp, Gyro, Acc)
Description
This system measures real-time ECG (electrocardiogram),
BCG (cardiotocography), and body temperature data using
an attachable bio-signal patch. and measures blood
pressure noninvasively using NIBP (noninvasive blood
pressure). The time difference between the R peak of the
ECG and the J peak of the BCG (R-J interval) is analyzed to
derive a prediction equation. In the future, machine
learning models will be integrated to improve the
prediction accuracy.
Tele Haptic Auto
Recognition Model
Customer : ERTL
Software : Machinen Learning, Sensor Data measurements
& Actuator
Description
- Hcube developed a Tele Haptic system upon request of ETRI
- By measuring the sensor data developed by ETRI,
classifying the touch through the machine learning model, and operating a small actuator according to the classified results to make users feel different vibrations
- The technology will continue to be studied in the future
Wire Bonder Predictive
Maintenance System
Customer : Semiconductor Test House Company
Software :
Description
- Wire bonder, an essential equipment in the semiconductor
field, currently uses more than 90% of foreign equipment, and periodic parts replacement is also costly
- After accumulating learning data through the data
measurement system, we are developing a system that aims to reduce the maintenance cost of the system by developing the most suitable machine learning model through analysis to identify the performance degradation outcome of the equipment
Solenoid Valve Controller Diagnosys
Customer : KIMM, Automotive Valve Company Joint Task
Software : Machine Learning, Auto Test Program
Description
- The vehicle solenoid valve is the most important part
because it is used for driving and stopping the vehicle
- Hcube has developed a system that not only determines a
simple defect in the solenoid valve controller but also infers and diagnoses the type of product defect
Machine Learning
Data Acq System
Customer : BAT
Software : LabVIEW, Python
Description
- BAT is a world-renowned tobacco producer, building a
machine learning
- based product management and product line maintenance
system
- They wanted to solve the problem of production stoppage
due to frequent breakdowns in the production line
- Hcubed provides DAQ and Machine Learning System for
Condition Based Maintenance
