Ron Gordon
Director » Power Systems
I recently attended a briefing and demonstration at the Arrow Electronics Open Lab in Denver, Colorado—a technology playground co-sponsored by business partners Arrow Electronics and IBM that showcases the potential and application of the Internet of Things (IoT) and gives customers the opportunity to explore the capabilities of technology to improve their operations, sometimes called Operational Technology (OT). There were about 50 people at the event and everyone was most impressed and walked away with a lot more knowledge about the capabilities of IoT and how to bring them to reality. We moved through eight different stations with fascinating use cases that really opened our eyes to how pervasive IoT is already and the wide range of capabilities that lay before us.
Welcome to the Arrow Electronics Open Lab
The 9,500 square foot facility, comprised of briefing rooms, multiple engineering laboratory settings, and several IBM OpenPOWER LC and IBM Power System AC922 servers is constructed so customers can interactively learn all about IoT solutions and how they can be applied at their company. Proof of concept (POC) environments are available where developers can, in a relatively short period of time, follow workflows to create a prototype that matches their business needs. The solution normally involves an instrument with a Bluetooth Low Energy (BLE) IoT device with sensors, for example a vibration monitor, a visual scanner, or a transmitting thermometer. (Just Google BLE devices and you’ll see the multitude of options that can support an IoT solution!)
The BLE device connects to a router which transmits the data to an analysis system where appropriate actions can be determined. Depending on the business needs, data may be logged for reporting purposes or, if determined to be out of specification, it can trigger an alert.
What Part Does Artificial Intelligence Play?
Artificial Intelligence (AI) and machine learning are used in IoT solutions to develop a model to analyze data in real time. IBM manufactures multiple solutions that support model development, including Power AI Vision, IBM Watson Machine Learning Accelerator, and IBM Cloud Pak for Data. The speed is incomparable. When developed on the IBM Power System AC922 with the integrated Nvidia GPUs and NVlink, AI and IoT can be put into practice far faster than with x86 solutions. The model execution can then run on the lower-cost IC922s with the data transmitted from the instrumentation. In cases which primarily call for data gathering and reporting, e.g., for regulatory needs like hospital refrigeration tracking and reporting, this simple function can be implemented by having the network access IBM Watson IoT in the IBM Cloud, which is low-cost and requires very little setup or management.
With IoT, AI uses real-time data and real-time actions, whereas other AI applications work on data that has been gathered but needs a preparation phase before it can be analyzed for prediction of what actions should or can be taken.
Collaboration for Operational Technology
Defining the IoT solution use case is the first step and can usually be defined and specified by working with people in operations. These may include the plant manager, a quality control manager, or a chief engineer. Use cases that are presented and discussed at the Arrow Electronics Open Lab in Denver, some of which run on the IBM Watson IoT Platform, include a variety of scenarios:
- On a manufacturing line, a model was created to determine “good” and “bad” quality points by using an iPhone for visual capture and inputting the image to the IoT process over the production line. As items go by, their “scanned” images are tested against the model by an IBM Watson Inspection application on the iPhone and alerts are generated to take action on the poor-quality items.
- In an agricultural environment, drones with a thermal imaging camera fly a pattern over fields and record the heat prints of various plants to determine growth, invasive species, and soil temperatures. This information is then transmitted to the model running on an IBM Power System server so the data can be analyzed and actions can be taken to optimize harvests.
- In a plant environment (HVAC, power utilities, manufacturing, etc.) the placement of vibration sensors on each motor assists in maintenance and trouble spotting before a motor breaks down.
All of these cases, and many more, can save on labor costs, improve accuracy, and improve quality.
The global market for IoT solutions is on a trajectory to hit $1 trillion in the next couple of years. If you’re interested in being a part of that, reach out to your Mainline representative. We can put you in contact with the Arrow Electronics Open Lab in Denver to set up an appointment.
Also, many of these topics will be presented at the IBM Power Technical University May 18-22 in Orlando, Florida and October 5-9 in Atlantic City, New Jersey. We look forward to seeing you there!
Please contact your Mainline Account Executive directly, or click here to contact us with any questions..
Related Information: