MicroAI™ and Asset Performance Management – A Specific Application
post-template-default,single,single-post,postid-20861,single-format-standard,bridge-core-1.0.4,translatepress-en_US,ajax_fade,page_not_loaded,,qode-content-sidebar-responsive,qode-theme-ver-18.0.6,qode-theme-onetech,wpb-js-composer js-comp-ver-5.7,vc_responsive

MicroAI™ and Asset Performance Management – A Specific Application

MicroAI™ and Asset Performance Management – A Specific Application

22 Jul 20


MicroAI™ and Asset Performance Management – A Specific Application

In this Article, we will discuss a use case for Asset Performance Management (APM) using MicroAI™. It introduces a simple concept that can be applied to any machine. At first, we will need to define our case and end goal. We will then list all the components needed to achieve that goal. Later, we can implement this and run test cases. Let us start out with defining our use case.

The APM Objectives and Requirements

We need to monitor a robot arm. We need alerts to be generated when there is a collision, abnormal increase in temperature, and abnormal sound pattern. All these cases will be our building blocks for our use case. This type of monitoring brings a lot of value to any business that depends on reliable asset performance.

Now, let’s list out things that we need. We need a prebuilt and configured robot arm that moves from left to right. We also need Raspberry Pi to run MicroAI(TM). We will need three different type of sensors: a sound sensor, a temperature sensor, and a gyro sensor.

We need to model the MicroAI™ before executing it. Before doing that, we need to mount the sensors, remembering that the position of sensors does matter. This is especially true for the gyro and acceleration sensor. If you position it incorrectly you will get negative readings. In this case, we will mount the gyro and acceleration sensor on top of robot arm tip. This way it will monitor the movement of the robot arm precisely. Sound and temperature sensors will be mounted on top of the robot’s motor housing. After mounting, make sure the sensor’s readings are accurate and make any necessary changes.


Implementing MicroAI™ into the Use Case

Now we are ready to execute APM using MicroAI™ on our robot. There are a couple of things that we need to keep in mind. It is always a good idea to model MicroAI™ for longer periods time. Also, how you use the MicroAI™ output can vary as well. For example, one can program to send an alert on an email, text message. or just a simple pop up stating alert.

At this point, everything should be running in live mode. We have our robot arm moving from left to right and vice versa. We have all the sensors connected on the robot to monitor it. After running it for about 10 to 15mins we can run a test to see if we can create an alert scenario. For example, introducing loud noises or abnormal levels of temperature to trigger an alarm.

Coming back to collision detection, we can try to stop or push the robot arm in an opposite direction when it is trying to move in its normal pattern. Be very careful as some robot arms can take up to a certain amount of tension before they stop working. If done correctly we will see an alert generated from our gyro and acceleration sensor.

Keep in mind, that this is just one simple use case for APM using MicroAI™. There are many other possibilities for integrating MicroAI™ into simple or complex APM use cases.