Operations Control Systems
Do Your Control Systems Lead to Profitability & Growth?
Once plant layout is optimized into continuous physical flow (as explored in Cellular Manufacturing), management faces a new challenge: maintaining operational visibility without creating administrative bloat. This brings us to the third pillar of cost architecture: Information and Control Systems.
Control systems seek to reduce the disparity between what is happening and what should be happening. These systems have evolved from personal observation of the master artisan and local verbal instructions in the pre-industrial era to the autonomous and predictive digital systems of today.
During the evolution the systems passed through the mechanical scientific management, psychological motivational focus and data-centric digital processes.
In tiny and very small organizations the pre-industrial era approach could still be valid. However the common practice of imposing cutting edge tech over an inherently broken process never works. Automating a mess only automates the mess; it does not solve the original problem.
Case Study of Automating a Mess
Fast Fashion Retailers had a messy inventory "control" system with disconnected spreadsheets and manual warehouse tagging. The inventory was moved "forcibly" into stores based on flawed historical forecasts.
Instead of fixing the inventory tracking and sales forecasting systems, they invested $15 million in an AI driven predictive replenishment system paired with an automated warehouse sorting system. The expectation was that the AI would automatically read inventory levels and instruct robots to pack and ship goods instantly.
The AI analyzed spreadsheet data containing an error in a manual log. The erroneous data led the AI to predict a massive surge in demand for a specific winter coat. The system instantly triggered high-speed production orders.
Physically, a human picker would have noticed a box that was misplaced or that coats were piling up. The digital system, operating at high speed, relentlessly pushed thousands of incorrect items into the automated shipping lanes, causing a total physical gridlock in less than 48 hours.
An additional problem was that the automated sorting robots relied on physical barcodes being perfectly placed on boxes. Because warehouse staff were never trained on standardized physical labeling, the robots misread thousands of smeared or misplaced labels.
Result:
Inventory error rate climbed to 45% from the earlier 12% of the manual system
Order fulfillment time rose to 18 days (owing to the warehouse gridlock) from 4 days under the manual system
The company lost $8 million in scrapped tech and cancelled orders compared to the moderate margin losses earlier
It now had a completely broken system that generated errors faster compared to the earlier fragile and slow manual system
Lesson:
Technology is a MULTIPLIER, it multiplies not only successes but also failures. Clean up the processes first before tapping technology.
Control Systems Have a Human Element
Modern sensors can sense variations in speed, temperature, volume, voltage and more, and generate alerts when the variations exceed permissible tolerances. Modern technology also enables sending these alerts to the offices and devices of decision makers so that timely remedial actions can be taken. It is also possible to use AI to automate the remedial actions.
However, even now,the effectiveness of these can be enhanced by tapping the human element, as demonstrated by the simple Shadow and Andon Boards.
Shadow Boards
The original shadow boards had images of all the tools needed at a specific workstation. If an image is not covered by the relevant tool (and the tool is not presently in the hands of the operator) it means the tool was missing. In that case, the operator can take immediate action before starting work.
Shadow boards helped ensure that the worker never had to go searching for a tool while busy with regular operations.
Andon Boards
Andon systems are visual systems that typically use colored signals to display the live status at a work station. Green meant work was going on normally, yellow indicated the need for assistance while red signalled stoppage of work.
In addition to displaying status as above, Andon systems also empowered workers to stop work when a relevant problem was noticed. This was in contrast to the traditional practice of only supervisors or managers having the authority to stop the work.
In the traditional environment, if the decision maker was not available, or did not attend to the problem, work proceeded with the faulty output moving as input to the next process.
When the faulty product was ultimately flagged at the end, it would have undergone many processes and the repair work (typically done at a specialized rework station) costs many times what it would have cost at the point of occurrence by stopping the work there.
Modern Shadow and Andon Boards
The digital revolution has modified these boards too, and today one will find digital equivalents doing the same work, viz. signaling problems in live time and empowering workers to stop work when critical failures occur.
Both Shadow and Andon Boards were pioneered by the Japanese car maker Toyota as part of lean manufacturing. As the world's largest manufacturing hub, China has adopted these extensively. Large automotive, electronics and multinationals in China invariably have these systems in place.
However, in smaller manufacturing units, these are often for show, to impress external auditors. You will find Shadow boards with missing tools and Andon signals being ignored.
Conclusion
The goal of control systems is to align actual performance with desired performance. However, systems by themselves cannot correct inherently flawed operating processes. In fact, imposing advanced tech on such flawed processes can speed up the errors.
Modern sensors and telemetry have made it possible to detect and remedy operational problems in real time. However, these need actual human involvement for effective implementation.
Shadow and Andon boards are great examples of simple options for effective control. These also demonstrate how the human element can enhance control.
However, even these can be just fancy showpieces rather than real control devices if the humans don't care.