< Back to Data & Analytics - Insight

Building on the AI/ML Foundation with Imagery and Predictive Maintenance

From the days of multiple Hadoop distributors battling it out for market share with everyone in a mad scramble to create larger and larger data lakes to the consolidation of that space into arguably a single solution provider which has worked arduously to remain a contender in the data ecosystem – there is no question about the incredible speed in which the data and analytics space continues to evolve. This evolution has forced many organizations to refine (and sometimes redefine) their data and analytics strategy to ensure that the tools, solutions, and processes are in step with the business’s needs. New and innovative solutions and means of automating processes have created opportunities to use data in new ways and collect through various imagery types while analyzing massive volumes of data at a staggering pace. Below we uncover how these modern solutions have positively impacted the manufacturing industry.

Image Analytics

Image analytics have seen tremendous advances as organizations across the globe leverage this process to extract meaningful information from images through digital imaging processing techniques. This helps improve operational efficiency, enhance near real-time decision-making, increase employee productivity, reduce cost, and limit potential employee physical harm. For example, let’s look at an analog meter or an LCD display that a system can read, decipher, and use in calculations. There are many instances in a manufacturing environment where the operators read displays on machines that don’t have an automated capability to send and receive information and make decisions regarding that machine’s operation. Enter the use of both artificial intelligence (AI)/machine learning (ML) and imagery. We now have a “lens” to read the analog or digital displays and make the necessary decisions to keep that machine in operation, helping keep the overall equipment effectiveness (OEE) as high as possible.

Lasers and Thermal Imagery

While the human eye can detect many imperfections, the use of lasers and thermal imagery can see even more. We can use acceptable standards to ‘look at’ the manufactured items to determine if they meet the appropriate standards. Rejecting substandard items before being sold or distributed reduces costs associated with returns and negative customer satisfaction (a single unfavorable tweet can cost thousands of dollars to remediate). Consider the use of imaging in the manufacturing of automobiles. Many of the auto plant tasks are now automated, especially those considered dangerous, such as welding. Now, a robot completes the frame’s welds, which are then inspected via standard and thermal imaging to ensure that the frame has been appropriately constructed and meets all safety specifications.

Predictive Maintenance and Imagery Types

Now that we understand how AI/ML with imaging makes manufacturing more cost-effective, we can introduce the idea of Predictive Maintenance (PdM). We know that machines need maintenance to continue to operate and that it costs less to maintain that machine than it does to fix it once it breaks. Imaging can visually (or thermally) inspect various items to determine if some intervention is necessary before a failure occurs. Our most recent development of an IoT and Predictive solution helps a large industrial client quickly analyze part specific information (service life, material characteristics, etc.). Utilizing image analytics to enhance visual inspections of equipment, improve the inspector experience, and enhance the streaming data’s analytic value.

Various industries and areas can leverage different types of imaging. There are many circumstances where inspection would be unsafe for humans, such as the interior of certain sections of a nuclear power plant or the stern of an oil tanker while it is still at sea. In such circumstances, drones can play an integral part. Consider agriculture; drones can inspect large acreages for damage after a storm or blight caused by insects. As the number of resources working at a farm becomes scarcer due to the pandemic, this type of imaging can help save the crop.

The Marriage Between AI/ML and Imaging Systems

These advances are happening at a more rapid pace than ever before. Given our current circumstances due to the pandemic, the need for this type of evolution has never been more imperative. These innovations improve our lives’ quality by reducing faulty materials, increasing the useful life of certain items through an enhanced PdM, and replacing the reductions in the availability of human resources. The marriage between AI/ML and imaging systems supports the economy and the population as a whole, and the benefits of such technology will last long beyond the pandemic.

Contact us to learn more about how Ness can help modernize how you collect and analyze data using imagery.

– Scott Schlesinger, Senior Vice President – Global Data Practice Leader
– Aaron Gavzy, Senior Data Strategist