4 Stages to Fully Automate a Concrete Block Factory

The ultimate dream of every factory owner is to run their operations fully autonomously, without reliance on human labor or the potential for human error. Although achieving complete automation is not yet feasible, the concrete block manufacturing industry can already take significant steps in this direction. The initial phases of this journey are achievable using current AI and robotics technologies, while further advancements will require substantial R&D.

The core approach to building an autopilot system is data-driven. It is critically important to collect and monitor three key types of data in real-time during the production process: factory and technological settings, the current state of the manufacturing process, and the parameters of the final product. Only the combination of these three data sets allows the AI core to provide recommendations or even adjust production settings autonomously. Another crucial component of an autopilot system is the set of actuators, which must be controlled by the AI to perform tasks that are currently handled by humans.

This paper outlines several stages toward fully automating a modern semidry vibropress factory operating across three shifts. Each stage will describe the necessary technological improvements and the benefits they bring. The article is intended for readers familiar with this industry, avoiding unnecessary explanations of standard terms and processes.

Stage 0: Current Industry Standard

In a typical factory, each shift consists of at least one operator and two sorters. The operator must be skilled enough to manage all factory operations and remain on duty

throughout the shift. Sorters are responsible for distinguishing defective products from good ones and must stay vigilant to avoid errors. Only during downtime they assist with loading packing machines and cleaning.

Summary for this stage

● Number of workers: 9 (operator and two sorters per shift)
● Human factor impact: High
● Additional equipment cost: None

Stage 1: AI-Based Computer Vision System

The first step towards automation involves installing an AI powered computer vision (CV) system capable of inspecting molded products on the wet side of production. An integrated weighing system, placed immediately after the press, measures the weight of product boards, enabling the calculation of average product density.

With this system, the operator’s attention can be significantly reduced, as any anomalies in the production process will trigger alerts. This allows the operator to focus on other tasks during the shift. A similar effect can be achieved on the dry

side by sending defect information through the tracking system. This will reduce the overall attention needed and lower the qualification requirements for the sorters.

Summary for this stage

● Number of workers: 9
● Human factor impact: Medium
● Additional equipment:
-Wet-side CV system (QuatroPBC)
-Weighing system
-RFID tracking system

Stage 2: Semi-Automated Production Process Control

The next phase of automation includes additional machinery aimed at increasing productivity. In this stage, several critical systems are installed:

● An RFID tracking system to track products from the wet side to the dry side,
● An additional CV system on the dry side to manage sorting process (QuatroPBC/DS),
● A product sorting system to replace defective products with good ones.
● A board dumping system to remove all concrete from the board in case of drops of excessive amounts of defective products.

However, at this stage, human oversight is still required. New position should be created to supervise the system's decisions about the quality - quality control specialist. He must partly or fully review the product batches processed by the AI to ensure the right decisions have been made by the system. It is especially important during early stages when AI decision-making is not fully reliable.

The specialist’s role will be to review AI decisions, using visual inspection tools, which takes approximately four hours per day to examine the whole 24-hour factory working day. The examination process will be done for the products in the curing chamber and will not result in any production delays.

The second CV system installed on the dry side helps handle any defects that occur between the wet and dry sides and provides accurate product coordinates for the sorting system. Since the probability of new defects between the wet and dry sides is low, all discrepancy data can be checked and confirmed on the dry side in real time by the operator.

Summary for this stage

● Number of workers: 4 (operator per shift and quality control specialist)
● Human factor impact: Medium
● Additional equipment:

-Dry-side CV system (QuatroPBC/DS)
-Board dumping system
-Product sorting system

Stage 3: Fully Autonomous Production Process Control

Once sufficient data is gathered, the AI system will be trained to operate autonomously, eliminating the need for a dedicated quality control manager. Periodic checks can still be performed by a parttime worker, or even remotely by an outstaff person as a service. Continuous manual oversight will no longer be necessary and the factory will be able to operate with minimal human intervention.

Summary for this stage

● Number of workers: 3
● Human factor impact: Low

Stage 4: Full Automation with AI-Driven Mix and Vibration Sensors

The final challenge in fully automating a concrete block factory lies in automating the control of concrete mix conditions and the vibration process. Currently, human oversight is required to adjust for variations in material consistency or excessive moisture. Similarly, the vibration process, crucial for production, is influenced by various factors, such as machine settings, mold condition, and technical

aspects of the equipment.

However, with future advancements in sensor technology, systems capable of monitoring and adjusting both the mix quality and vibration settings in real-time will likely be developed. Once these systems are integrated with the AI core, the factory will be capable of fully autonomous operation.

Summary for this stage

● Number of workers: 0
● Human factor impact: None
● Additional equipment:

-Concrete Mix Control system (R&D required)
-Vibration Control system (R&D required)

Conclusion

Achieving full automation in a concrete block factory requires the installation or development of several key systems:

● A board weighing system to calculate density, a critical parameter for concrete;
● An RFID tracking system (or equivalent) to monitor products at various stages;
● A computer vision system to inspect products on both the wet and dry sides and measure their parameters;
● Sorting and dumping systems to handle defective products;
● A concrete mix control system for continuous monitoring of the mix’s condition (R&D required);
● A vibration control system to measure the actual impact of vibrations on the mix within the mold (R&D required).

It is still impossible to implement all these systems in a factory at once due to their

absence or insufficient tradition of application of some of the systems. Quatromatic proposes to consider the path to full automation as a step-by-step process of 4 steps, which should be overcome sequentially.

It is already entirely feasible to implement Stage 1 in any factory today with the

installation of an appropriate CV device. Leveraging the expertise of modern machinery manufacturers, integrating the CV device with other production line equipment to achieve Stage 2 is a straightforward process. Stage 3 will soon become a reality for concrete block factories that are ready to embrace this innovative approach to the production process by changing some roles for involved employees. Stage 4 presents more challenges, as it requires extensive R&D in two distinct areas, but the successful implementation of the earlier stages is likely to prompt global leaders in concrete mix design control and vibration control to address these issues in greater depth.