Work Like a Machine (Tool)

June 20, 2018
An ode to real time analytics

In a display of the absolute absurdity of this data age, every Monday my computer sends me an email about myself. Specifically, it provides me with a sort of roundup of all the ways I spent my time during the preceding week.

Maybe it’s because I’m a geek, maybe it’s because I devote so much time worrying about data and analytics, or maybe just because I love absurdity, but I’m totally obsessed with these weekly emails. Not so much for their content, but for the broken sense of analytics they represent.

According to these reports, I spend roughly 60% of my office time in meetings preparing for the work to be done, and another 15% of my time sending and reading emails about the work to be done. That leaves me with just 25% of my week—not even one full day—to actually do the work. This is why, I’m sure, the system always scolds me for working too much after hours.

There’s probably nothing special about these results; I imagine they’re roughly the same for any other desk jockey out there. But that’s hardly an excuse, and it sure doesn’t make it right.

For context, I like to translate this into more industrial terms—into, say, machine tool utilization. Because I’m here to do work and create something, after all. Just like anyone else.

In that light, this weekly email would send alarm bells ringing all the way up through management. Just imagine the impact of 75% downtime on a machine tool, or operators spending ¾ of their day preparing for a job, and then working overtime to get it done. Bosses would lose their minds, people would get fired, contracts would be lost. It’s fun as hell to imagine when I see how I spent my life.

Here’s the kicker, though. In this scenario, the report is coming out a week later. It contains all kinds of impressive data, broken down into helpful buckets of time allocations. But by the time the report is issued, the damage is already done.

This is the real issue here, and why the urgency in the industry today goes so much further than just data and reports.

Data, as I’ve written here before, is dumb. It takes no account for context, no account for the complicated systems and realities of its content. Which is why we focus so strongly on analytics.

In our my machine tool example here, knowing my uptime does no good unless we can place that data against the jobs I’m supposed to be processing. Even better, placing it against the full schedule of jobs coming after it.

But even that insight would be neutered by a delayed report. Knowing that I’ve missed a deadline after I’ve missed it is just as good as not knowing at all. Knowing why I missed it doesn’t really add much value.

No, the real objective in this long walk of data integration isn’t to know what or why something happened, but to know what’s going to happen—and what we can do about it.

That is the real power demonstrated by MachineMetrics in our cover story for the June issue of New Equipment Digest. It’s pulling all of this data and all of these reports together in context to provide a real-time indication of the true state of the world—of where we are in our process, about what that means to future processes, and what we need to do right now to keep everything running smooth.

It’s a beautiful thing to see come together. It’s exactly what we need from our machines, exactly what we need in this big wide world of big data. And eventually, maybe, we can incorporate some of it into our work lives, too.

But for now, that might be dreaming too big. We humans are still stuck in reports, still stuck to late data, still stuck in meetings. I’m late to one now. But I’m sure my computer will remind me of that next week, long after it’s forgotten.