AI Apps for Manufacturing Strategy Council
Oracle Modern Business Experience (MBX) is an Oracle conference that was held March 19-21 in Las Vegas. Unlike Oracle Open World in San Francisco which covers database, technology, and applications, MBX is solely focused on financial, supply chain, and human resource software applications that is needed to run global and local enterprises. At MBX, the AI Apps for Manufacturing Strategy Council was held at 9AM on Tuesday. Despite the early start, the room was packed by 9:30AM, and was standing room soon afterwards for nearly the entire session. A scan of attendees reveal that over 25 companies from diverse industries including:
- high technology
- renewal energies
- industrial equipment
- steel processing and manufacturing
- electronic components
- building material
- consumer goods
Manufacturing Latching on to Industry 4.0
Industry 4.0, aka Smart Manufacturing, Advanced Manufacturing, or Intelligent Manufacturing, is defined in this 2013 document from Germany’s National Academy of Science and Engineering. Fast forward to 2019, and that document might look quite different. Today’s version of Industry 4.0 might include these technologies that were either nascent or not even invented back in 2013: 1) A/R, V/R 2) Big Data 3) Mobile 4) Artificial Intelligence 5) Cloud 6) Computer Vision 7) 3D/4D Additive Printing 8) Digital Thread 9) Block Chain. At the AI Apps for Manufacturing Strategy Council, we surveyed the current level of Industry 4.0 technology adoption, and the results were mixed. Some already had implemented Industrial Internet of Things (IIoT) sensors, storage systems (Historian), some form of analytics (embedded or external), and some even implementing a mild form of “remote” or “lights out” manufacturing. But most are looking to build their next factory with Industry 4.0 as a key requirement.
AI and Machine Learning: The Next Focus for Industry 4.0
Whilst the adoption of components of Industry 4.0 is already underway, the adoption of AI and Machine Learning has not kept up with the adoption of PLC, MES, and IoT. Operational Technology (OT) data, the generic category name for IoT data, has been “easy” to collect. But much of that data is stuck in data siloes such as Historian or MES systems, where the data sits and rots. The path to extracting this data has not be a high priority, unmotivated by what these data can reveal. But AI and ML now is coming to the rescue. With these technologies, previous siloed OT data can be mined to reveal deep insights about manufacturing that were previously difficult.
Core Features in Oracle AI Apps for Manufacturing
Oracle AI Apps for Manufacturing is a machine learning powered analytics application tuned specifically for manufacturing. The application sits on top of a built-in datalake. This datalake was designed to connect to heterogeneous systems: both live Operational Technology (OT) and business centric Information Technology (IT). What makes Oracle’s application unique is that it is specifically tailored for manufacturing. So typical manufacturing data schema is already mapped into templates, making adopting AI for manufacturing much easier. Oracle AI Apps for Manufacturing has four major features that uniquely enables manufacturing organizations to monitor, control, and impact manufacturing.
- Insights: Explore machine learning driven patterns and correlations from historical manufacturing big data that affect operation efficiencies.
- Factory Command Center: 360 degree view of operations across manpower, machines, materials, methods, and management.
- Genealogy & Trace: View backward and forward trace of products from disparate operations and informational technology systems spanning manufacturing and supply chain processes.
- Predictions: Review predictive alerts and influencing factors about operational metrics such as yield, defects, rework, scape, cycle time, costs, etc – to enable timely corrective actions.
Strawberry Jam Factory Live Demonstration
At the Strategy Council, a live demo of AI Apps for Manufacturing on a strawberry jam manufacturing was shown. Jam consistency is one of the main qualities consumers expect in a jam. But multiple factors impact the consistency of jam. In the demo, we showed that:
- Factory Command Center: gives you a real time view of your jam manufacturing, using the 5M framework across manpower, machines, materials, methods, and management.
- Genealogy & Trace: a bad batch of strawberries is easily traced upstream (to the supplier) and downstream (to stores that has jam made from the bad batch of strawberries).
- Predictions: an oven is starting to exceed the upper limit of operating temperature range, which will produce sticky jam consistency; an alert is generated to fix the oven BEFORE the bad jam is produced
- Insights: contextualize OT data with IT data, then find deep patterns and correlations (such as a combination of operator, machine, and time) that contribute to low output or bad yield
In a standing room only session, the presence of diverse types of companies and the varied depth of adoption of Industry 4.0 has shown that AI and Machine Learning is the next big thing for manufacturing. For more information on Oracle AI Apps for Manufacturing, please visit https://cloud.oracle.com/ai-apps-for-manufacturing