Using Industrial Automation to Improve Productivity

Throughput is lost in small pieces

When people think of automation raising productivity, they often picture a machine working faster. In practice, much of the available gain comes from removing small losses that add up: a few minutes waiting for material, short jams that need manual clearing, a slow restart after a break, and a steady trickle of defective parts that must be reworked. Automation that targets these losses can lift throughput without any machine running at a higher top speed.

A hypothetical shift

The figures below are invented for illustration. Check the arithmetic with your own data before drawing conclusions.

Imagine a line with a cycle time of 30 seconds per unit. A shift lasts 8 hours, which is 480 minutes. Due to waiting and short stops, the line sits idle for 60 minutes per shift, so it runs for 420 minutes.

• Running time in seconds: 420 times 60 is 25,200 seconds.
• Units made: 25,200 divided by 30 is 840 units.
• Suppose 5 in every 100 units need rework. That is 42 units (840 times 0.05).
• First-pass good units: 840 minus 42, or 798.

Now suppose automation cuts idle time to 30 minutes by feeding material automatically and clearing minor faults faster, and improves a fastening step so that only 2 in 100 need rework.

• Running time: 450 minutes, which is 27,000 seconds.
• Units made: 27,000 divided by 30 is 900 units.
• Rework: 900 times 0.02 is 18 units.
• First-pass good units: 900 minus 18, or 882.

First-pass good output rose from 798 to 882, which is 84 more units per shift, and the machine’s cycle time never changed. Note also that the rework labor fell from 42 units to 18. The example is deliberately simple; real plants face variable demand, product mix and other constraints.

Practical ways to cut idle time

Automatic material supply. Feeders, hoppers and conveyors that deliver parts when needed keep a machine from waiting for an operator. Conveyor throughput and the flow of material between stations are frequent sources of hidden delay.

Fault detection and quick recovery. Sensors that identify a jam or a missing part, and controllers that respond with a defined recovery routine, shorten stops. A clear message on an operator screen that says what happened and where helps people fix it faster than a generic alarm.

Automatic changeovers and recipes. Storing product settings in software reduces the time and error involved in switching products.

Coordinated start-up and shutdown. Sequences that bring a line up in the right order, and restart it after a stop, reduce the idle minutes that occur at the start of shifts and after breaks.

Visible reasons for stops. Even basic automatic logging of when and why a machine stops often reveals that a small number of causes account for most of the lost time.

Practical ways to cut rework

Check at the source. Presence sensors, simple vision checks and torque or force monitoring verify that a step was done correctly while the part is still in the machine.

Error-proofing. Fixtures and sensors that make it impossible to assemble a part the wrong way prevent defects rather than detecting them.

Closed-loop adjustment. Where a measured value drifts, a controller that corrects it automatically keeps output within tolerance.

Traceability. Recording which settings and materials were used for each batch makes it easier to track down the cause of a defect. A root cause analysis is then based on facts rather than memory.

Choosing where to start

Not every idea is worth pursuing. A reasonable sorting method is to estimate for each candidate:

1. how many minutes per shift it could recover, or how many units of rework it could prevent
2. how reliably you can measure that today
3. the likely cost, including installation, training and maintenance
4. the risk, including safety and the effect on production during installation

Start with items where the loss is clearly measured, the fix is simple and the risk is low. Some of these may not need much automation at all. A well-placed sensor and signal light, for example, is inexpensive compared with a robotic cell.

For a small workshop, the principles are the same on a smaller scale, and the ideas behind small business automation apply.

Watch for side effects

• Bottlenecks move. Gains at one station may expose the next limit.
• More output needs more input. Faster flow requires material, packing and shipping to keep up.
• Maintenance becomes more important. Fewer stops per day means each remaining stop matters more, and automated systems need regular care.
• Procedures and safety. New equipment changes how people work near machines. Use trained personnel, follow manufacturer manuals and lockout practices, and consider local regulations.

Keeping the gain

Record a baseline before changing anything, measure again after a settling period, and keep tracking. Gains that are not watched have a way of slipping.

This is general information; the numbers above are hypothetical and no results are promised.