Two manufacturing-focused solutions for AI-powered data analytics and equipment connectivity to enable Industry 4.0.
Google has two new solutions to enable Industry 4.0 for manufacturing facilities: Manufacturing Data Engine and Manufacturing Connect. Together, they help link equipment on the manufacturing floor to the cloud and facilitate advanced data analytics powered by AI.
The solution was designed to make advanced analytics a self-service tool that is accessible to everyone. According to Google, anyone involved in the manufacturing process can interpret the data output, not just technical IT experts.
“The goal of the solution was to provide a highly scalable and repeatable solution for manufacturers. Making it easier to create and onboard new factories using the same solution,” explained Simon Floyd, industry director, Manufacturing & Transportation, Google Cloud.
One of the solution’s key features is a low- to no-code setup and user platform. This reduces the deployment time and makes it ideal for a wide variety of end users.
For Industry 4.0, Data Is Not the Problem
In conversation with engineering.com, Floyd highlighted the fact that most companies are not struggling to generate data. “Manufacturers have data,” he said. “The challenge is contextualization and learning what to take from that data. Companies need to learn how to use data to inform their decision making.”
A single machine can produce 5 gigabytes of data per week, and a smart factory can generate 5 petabytes per week. Yet, with such massive amounts of data, most companies struggle with implementing analytics that provide real-time improvements in their production.
Not only are companies already collecting data, but many have also begun to integrate cloud-based data solutions. According to a 2021 McKinsey & Company survey, two-thirds of manufacturers are already using cloud-based AI solutions. However, despite the widespread use of cloud computing, many companies struggle to gain insight from their data and scale their solutions. The same survey found that almost half of respondents view their cloud technology as “more (or much more) complex than initially expected.”
So, with ample data and disparate cloud-based solutions, companies are frustrated with finding ways to leverage their data and realize the potential of Industry 4.0. To help companies make the most of this wealth of data, Google released two new manufacturing-focused solutions.
AI with the Google Manufacturing Data Engine
The Manufacturing Data Engine solution is designed to analyze, contextualize and store data for large-scale manufacturing companies. Using a secure and low-cost connection between the edge and cloud, the platform can receive data from various machinery. That means diverse data, from telemetry to images, can feed into the AI engine. The solution then uses a common data model with factory-optimized data to provide important context for analytics.
Beyond storage, the key feature of the solution is the AI-driven analytics used to provide insights across manufacturing applications specific to factory KPIs and strategic initiatives. The manufacturing analytics and insights feature is a no-code, easy-to-use solution designed for manufacturing engineers and plant managers to create custom dashboards with the information they need to make decisions. New machines, setups and factories can then be added to facilitate scaling.
One key feature highlighted by Floyd is AI-enabled predictive maintenance. Of particular importance is the speed with which the solution can deliver maintenance analytics. Prebuilt machine learning models can help manufacturers deploy predictive maintenance within a few weeks from the deployment date, as opposed to several months with other solutions. Briefly, the software looks for inefficiencies or equipment issues that will prevent machinery from achieving its scheduled maintenance window. The goal is to mitigate unplanned downtime and take advantage of real-time, cloud-based monitoring. The models can then be continuously improved in collaboration with Google engineers.
In harmony with the predictive maintenance feature, the Data Engine also includes a real-time anomaly detection feature to provide alerts at the machine level in factories. Use cases of the solution include monitoring noise, temperature and other sensors in real time to detect anomalies and prevent issues before they arise. These are all important features for Industry 4.0 applications.
With the volatile supply chain issues experienced over the past two years, Floyd also highlighted the ability of predictive maintenance to support supply chain health. “These insights allow [manufacturers] to get ahead of the supply chain; because the supply chain is not affected by just products, but also the manufacturing equipment deteriorating over time,” explained Floyd.
Rapid Edge Computing with Manufacturing Connect
Manufacturing Connect is an edge platform codeveloped by Google and Litmus Automation. The solution includes a library of more than 250 machine protocols that allow connection with almost any factory equipment. The solution converts specialized machine-based data into datasets that can be processed by the new Data Engine solution. Connect also supports containerized workloads to facilitate low-latency data visualization and analytics on the edge. This provides manufacturers with the flexibility to use existing IT or work with all their data in Google Cloud.
Sensors act as a gateway to transmit machine-level information to the cloud and facilitate edge computing. On the edge, response times can be in milliseconds. If manufacturers don’t require that level of speed, analytics can also occur directly in the cloud for response times in minutes. A combination of both response speeds can also be used depending on the needs of each piece of equipment.
For manufacturers using multicloud or hybrid on-premises solutions, the software is also compatible with BigQuery Omni to facilitate data collection across cloud environments. Data from all locations can then be collected and analyzed with the Manufacturing Data Engine.
In terms of cybersecurity, Floyd noted that the two solutions are built on top of Google Cloud. So, enterprise-grade security is available for both connected devices on the edge and analytics in the cloud.
Ford Is Already Deploying Manufacturing Solutions at Scale
Currently, Ford is operating two of its manufacturing plants with both the Data Engine and Connect solutions. This includes over 100 pieces of manufacturing equipment and more than 25 million records per week.
“[Ford] was collecting data at scale but couldn’t provide analytic capabilities they wanted for everyone,” explained Floyd. “Contextualizing the data was important as not everyone could understand the analytics. Now, anyone can gain meaningful insight.”
When asked about implementing the new manufacturing solution, Jason Ryska, director of Manufacturing Technology Development at Ford Motor Company, said in a Google Cloud video, “First and foremost, there is a lot of data that’s available and a caution in this space that simply because you can collect data on everything, doesn’t mean that you should. So, we start with a very clear problem statement and develop competencies around that. Once you get the system and the platform and the tools in place, you can scale that to other problems.
“The growing amount of sensor data generated on our assembly lines creates an opportunity for smarter analytics around product quality, production efficiency, and equipment health monitoring, but it also means new data intake and management challenges,” added Ryska in a press release. “We’re gaining strong insights from the data that will help us implement predictive and preventive actions and continue to become even more efficient in our manufacturing plants.”
Ford and Google are now working together to optimize their solutions further to meet the needs of current car manufacturing, especially as the automaker switches to more electric-based vehicles and cars with increased Internet of Things (IoT) capabilities.
Google Provides Solutions for Industry 4.0
With its new solutions, Google looks to be using its legacy of AI expertise to help manufacturers derive value from the big data they are already collecting. Recognizing its blind spot in hardware, Google partnered with others to help scale sensors that will deliver the type of data needed for impactful AI. Although many manufacturers are interested in the potential of real-time AI insights to optimize production, it seems many companies are frustrated by scaling solutions across factories and ensuring that anyone can interpret data, not just IT experts. Google also seems to have delivered on this front, leaning on its user interface experience to create an aesthetic and easy-to-use platform for users with any level of expertise. What remains to be determined is if the fast deployment and end-user-focused system will provide the scalability most companies want without sacrificing the actual power of the data analytics.