AI + Embedded engineering
Valve Predictive Maintenance System
Industry
Entertainment
Time
12 month
Team
8 members
Type
Startup
Valve Predictive Maintenance System
Project idea
Development of a predictive valve maintenance system to increase the reliability of production systems, reduce the risk of accidents and cut maintenance costs.
Project challenges
Our customer is a valve manufacturing factory that makes valves used in systems with liquids and gases and installed in various industries.
Often, manifacturing performance depends on the reliability of valves and the timeliness of their maintenance. Considering that these valves are also used at critical facilities, for example, military submarines or gas and oil distribution enterprises, etc., it is critical that the condition of units be analyzed on time anda decision on timely replacement be nade, if necessary.
Results
Research, development and implementation of an integrated monitoring system based on the collection and analysis of data from ultrasound and vibration sensors.
Using AI, Machine Learning and Deep Learning algorithms, and with continuous learning of the system for recognition of different states of the nodes under investigation, we have achieved more than 93% accuracy in both emergency and state situations.
Project development stack
See the ways how the development team implemented requirements
Neural network
-
Tensor flow
-
Keras
Cloud services
-
Amazon Web Services
-
Microsoft Azure
Data transfer
-
TCP/IP
-
MQTT, MQTT-SN
Wireless Data transfer
-
NB IoT
-
Bluetooth
-
Wifi
-
GSM
Tell us about Your Project