The New Data Trap

New Data

Autonomous Haulage gives you more data than you've ever had before, but more data does not mean better decisions. For example, you will now know why and for how long a haultruck was stopped at an intersection. That's good information to have, right? Sure... as long as you don't fall into the New Data Trap.

The New Data Trap is when newly visible Key Performance Indicators (KPIs) drive local optimizations without improving system throughput. There have always been stoppages at intersections in the mine, you just didn't know how many. Suddenly you have the data, so you want to "fix" the problem.

Here's where the Theory of Constraints comes in.

The Theory of Constraints

The Theory of Constraints is basically de-bottlenecking. You improve system throughput by focusing systematically on single limiting factors.

1. Identify the constraint

   Confirm that the issue you want to fix is actually limiting throughput.

   - If intersection stoppages are not reducing production below shovel or crusher capacity, they are not the constraint.

2. Exploit the constraint

   Improve performance using minimal capital.

   - Better traffic rules, priority logic, or operational discipline.

3. Subordinate everything else

   The constraint gets priority over competing improvements.

   - If the intersection is the constraint, fix it before widening other roads.

4. Elevate the constraint

   Invest capital if necessary to remove the bottleneck.

5. Prevent inertia

   Stop improving once the constraint moves.

   - Over‑optimization creates waste: restart at step 1 frequently

An identified bottleneck is a good thing, as it helps you avoid spending money somewhere else that will not yield results. It helps control the entire system.

Take Smart Action

The New Data Trap is what you fall into when you chase new data without first following the Theory of Constraints. Here are some intelligent steps to take when you get new data:

1. Determine if this new KPI is actually the bottleneck.

2. If it is, determine what good looks like.

   • what good looks like isn't zero stoppages, it is the minimum stoppage required to maintain safety while meeting throughput targets.

3. Improve the bottleneck, but only up to the point where it is no longer the constraint

AHS data is powerful, but only if you let throughput, not visibility, guide your decisions.

References

Bowater, Mark. Crimes Against Mine Planning: Solving the Top 10 Pitfalls. AusFV Pty Ltd, 2022.
which referenced
"The Goal" by Eliyahu Goldratt