High-volume factories with legacy equipment are an afterthought for most technology providers. For us, it’s our focus. No more disorganized software platforms that require an entire IT team and multiple systems to deploy successfully. No more computing solutions meant only for PhD data scientists. We deliver end-to-end solutions for your key stakeholders.
Featured Industry: Steel Tubes
Upgrade your production line to 100% visual inspection. Identify burns, inconsistencies in galvanizing thickness, scratches, or leftover scale in real-time. Identify right away when your line speeds, welding settings, or galvanizing settings need to change. Cut your scrap in half and keep production stoppages to a minimum.
We developed our beta platform in partnership with the one to ONE Holdings companies Daiwa Steel Tube Industries in Japan and Daiwa Lance International in Vietnam. See our case study below detailing the solution implementation and benefits:
Metal Rolling Mills
Identify visual defects during billet heating, roughing, pickling, or rolling. Correlate defects in rolling or cooling to sensor data measured upstream. Generate rules and standards to scrap defective steel or pause production BEFORE steel is rolled.
Paper & Pulp
Paper is generally produced with continuous process manufacturing, starting with the raw material of wood followed by fiber separating, bleaching, and further refinements. Within these primary stages are multiple steps with data points attached to each operation, making it ideal for utilizing IndustrialML’s methods
- Paper is initially mixed with processed water and then stirred in a stainless steel vat, called a pulper, to make a fiber suspension.
- Then paper machines remove water from the solution, which is comprised of 1% fiber and 99% water.
- A wire or forming fabric removes more water and ensures that the fibers are correctly weaved together.
- The paper web is passed through a series of rollers that squeeze the water out of the pulp mat. This compresses the fibers, so they intertwine to form a thick, smooth sheet.
- Drying cylinders get rid of almost all of the water at this point, having warmed the paper to upwards of 100 degrees celsius.
- In total, the journey from start to finish across the production line can be around half a kilometer. At the end, the paper is rolled into an enormous reel where it can be unwound into smaller rolls.
Regarding industry challenges in paper production, the most significant concerns are moving the paper through the stages as quickly as possible and reducing errors. Although there tends to be a lot of scrap during production, it is not a major concern from a monetary standpoint. Still, it will become increasingly prevalent as the supply of raw materials decreases.
The paper industry certainly has a lot to gain from digital manufacturing to reduce costs and better use the ever-decreasing raw materials. Data can be applied to almost all production areas, as paper and pulp production collects this information readily, so there is plenty of scope for it to be used for analytics and automation.
Where IndustrialML comes in
- Computer vision helps to improve the inspection of raw material quality. Integrating data from multiple sensors and displaying them in real-time allows remote monitoring to be possible.
- Real-time alerting from the sensors reduces the downtime and amount of scrap produced in the production line. Paper manufacturing is a consumer application so reducing the chances of a bad batch is imperative.
- In producing reams of paper, the processes are different from a conveyor belt production line discussed above. In this kind of batch processing, it’s easier to spot errors because there isn’t the same continuous movement. Remote monitoring along with dimensional analysis provides comprehensive quality inspection.
Food & Beverage
Factories involved in meat manufacturing take fresh meat from the abattoirs and processing plants and use it to make products like pies, sausages, bacon, ham, and frozen meals. The raw meat can be processed in several ways such as curing, mincing, adding flavorings, and various other techniques to produce their products.
Meat processing is heavily dependent on labor, with workers accounting for 15% of company costs. This also depends on the type of production. For example, there may be many workers in an assembly line for curing and cutting into specific shapes. However, for marinating poultry, a huge vat could be used for a large batch of meat.
There are increasing labor shortages, high worker turnover, and wages are also rising in the industry. Furthermore, regulations to improve food safety, worker conditions, and animal welfare have also led to higher costs. However, despite these concerns, the meat processing market is relatively stable - between 2016 and 2020, the production volume of meat worldwide has increased from 317 million metric tons to 328 million metric tons.
Automation in the meat manufacturing industry is well underway. Machines in a chicken processing plant use laser scanners to produce a 3D image of the specific cut (Fillet, thigh, drumstick, etc.) Then the computer controls an ultra-high-speed, rotating blade that cuts the specifically weighted portions. These automated systems are less prone to errors, but they could incur a higher cost than errors made by humans.
Where IndustrialML comes in
- Remote monitoring of the machine in the above example’s conditions let a manager know how many cuts are made. Predictive machinery maintenance can indicate how many cuts are advisable at the current operating level before changes need to be made.
- IndustrialML has a seamless audio communication system for manufacturing that would be well-suited to meat processing factories. Alerts can be generated and sent to an operator’s headset to be notified immediately when there is an issue on the production floor. Having to spot a red light or check notifications on a phone for an alert could lead the operator to miss something crucial.
- Finally, a camera feed is beneficial for monitoring the consistency of meat products so that no expired materials make it to the end of the production line.
Monitor production during preparation and mixing, manufacturing of tire components, and the formation and curing of tires. Understand the relevant data at each process step (ply cutting, belt calendaring, tread and sidewall extruding, etc.) which leads to good or defective products. Know right away when your stored rubber has reached its shelf life, BEFORE it is used in production.