There is a version of this in nearly every workshop. A breaker runs alternate weeks under two operators, and over a few months somebody notices that tools come back in better condition from one of those weeks than the other. Same machine, same face, same rock, same shift length. The foreman has known it for a while and has probably mentioned it, and it has probably gone nowhere, because the only evidence is that a man who has been rebuilding hammers for twenty years thinks so.
That technique matters is the oldest observation in the trade and the least documented. Every experienced operator will tell you that some people wear tools out and some do not, and every experienced supervisor will tell you they can hear the difference from across a site. None of it has ever made it onto a report, which means it has stayed as folklore: true, widely believed, and completely unusable for making a decision that anybody has to justify. You cannot schedule maintenance around a feeling, and you certainly cannot have a conversation with a crew about one.
It stayed folklore because the instruments measured the wrong thing. An hour meter records that a machine was on. Hydraulic flow records that oil moved. Neither has any way of describing how the work was done, because technique is not a quantity of anything, it is a pattern in time. To an instrument counting duration, a careful operator and a careless one produce identical output, which is why the difference everyone could hear never appeared in any system anyone had bought.
Motion is different, because technique is expressed as movement and an accelerometer reads movement directly. Two things are derived from that signal for a breaker, and they turn out to describe most of what separates one operator from another. The first is the distribution of firing burst lengths: how long the tool stayed down in each detected burst. The second is the share of the period classified as blank firing, the tool cycling into air rather than into material. Both are properties of how somebody works rather than of how long they worked, and both are classifications drawn from the motion record rather than observations somebody made at the face.
The two signatures point at different problems, which is what makes them worth reading separately. An operator whose bursts run long is committing to a spot and staying with it: energy going into the same point past the moment the material was going to yield, heat building in the tool, and the wear pattern that follows from that. A high blank firing share points at a busy trigger: short cycles, plenty of them, and energy repeatedly travelling back through the pins and body instead of into rock. Both wear the machine out. They wear different parts of it out, they call for different conversations, and treating them as one number called technique loses the only useful thing about the measurement.
The comparison worth making is the narrow one. Two operators, the same machine, the same site, the same period. That removes almost everything that would otherwise explain the difference: material hardness, machine condition, tool age, hydraulic settings, weather. What is left is much more likely to be the person. Comparing operators across different sites or different machines is far weaker, and it is the kind of comparison that produces confident conclusions about people that fall apart the first time somebody points out that one of them was working sandstone. SolidAI Hub will put two assets, two operators, two sites or two shifts side by side, provided the operator and shift assignments were made in SolidTrack beforehand, and the constraint to hold in your head is that the more the two have in common, the more the difference means.
The obvious next step is a league table, and it is worth thinking carefully before taking it. Ranked lists of operators do two predictable things. Some people start working to the metric rather than to the job, which on a breaking site means breaking more cautiously than the material requires and losing more in time than the parts were ever worth. Others conclude that the system exists to build a case against them, and once a crew believes that, every subsequent number arrives pre argued. Neither outcome gets you longer tool life.
The more productive reading is upside down: the crew with the best numbers is the useful one, not the worst. Whatever they are doing can be taught, and they can usually explain it if somebody asks properly. In practice the answer is almost always about anticipation rather than attitude. The operator with short controlled bursts and little blank firing is reading the material and releasing as it starts to go. The operator with a long tail is waiting to be told by the tool falling through. That is a learnable habit, and it is a much easier conversation than one that begins with a ranking.
Two honest limits before you act on any of this. The first is that operator attribution is only as good as your shift and asset assignments. SolidTrack does not work out who was in the seat; it reads the assignment somebody made in advance, so if the roster is approximate, or two people share a machine within a shift without it being recorded, then the numbers belong to the machine and not to anybody in particular. The second is that material varies within a site as well as between sites, and a week spent on a stubborn part of the face will look like heavy technique. Give yourself several weeks and several rotations before you conclude anything about a person.
None of this is a new idea about how breakers wear out. Every part of it was already known by the people doing the work. What changed is that the thing everyone could hear from across the site is now written down, attached to a machine and a period, and can be put next to another week and compared. That is a small change technically and quite a large one in what it makes possible, because an argument about a wear rate is very different from an argument about who is careful.