🦉 10x curiosity — Issue #75 — Industry of the future — The Office
What does the future of work in the process industry look like. Lets explore some themes…
Looking over the next couple of weeks to explore some possible futures in process plants and how various technology solutions might come together to change how these places operate. There are so many exciting developments coming online to transform work as we know it. Over the coming weeks I plan to work through a number of seperate environments of the process industry work place exploring how the not to distant future may look. These environments include:
- The office
- The field
- Equipment
Wanting to try something different I thought I would also write this as a narrative rather than a factual report. We’ll see how this goes…!
Up and at em…! Going to work is now a decision to be made, the 7:30 production review is attended remotely by most people. Holographic technology has made this the norm allowing everyone to avoid peak hour traffic, both ground and air! The production reports and optimisation opportunities had been compiled by the AI for review and vetted by the remote work force team based around the world — utilising different time zones truly means the process is continuously monitored. One particularly tricky issue had been posted to the open crowd sourcing competition site with half a dozen promising options to resolve it identified. These will be run through the digital twin to be validated and allow the best solution to be selected. Given the complexity of the run it might take several minutes but there should be a clear direction by the end of the 20 minute meeting.
Also identified by the digital twin were several instruments that deviated from the predicted model performance and likely required calibration. One of these was playing up for the third time so a maintenance request to replace it had automatically been raised, the parts were being printed on the 3D printer to delivered as the maintenance team arrived. Seeing a gap in the operations schedule the automated scheduling application had already rearranged the resourcing requirements, and flagged Jeff at the autonomous crane base to review the access with the VR googles — tight but should squeeze in.
Also of interest were a number of alternate process setpoint options that the digital twin had run overnight, highlighting a 2% production improvement by adjusting 3 setpoints — two were straight forward but the third was a bit left field, one we never would have thought of before. Once implemented and validated with plant data the feedback would be stored in the extensive machine learning library and more rapidly accessed the next time this process scenario arises.
With the preliminary optimisation complete it is time to dive into some more detailed analysis of a longer term problem that is creating a continual bottleneck for the process. Starting with a literature search of the extensive plant archives — A few key words links to likely memo’s, a comprehensive capture of all relevant photos in the system and the searchable log aggregates the keyword mentions and converts the log text in to a time series text analysis. A new data source beyond just numbers. Video logs also exist for all equipment back several years and it is only a matter of minutes to overlay video files of the equipment during these poor periods of performance to begin to dig into what is happening.
Processing so much data is only now possible thanks the the analytics upskilling the team has gone through. Using PowerBI, R and Python combined with the existing stats skills most process engineers developed out of uni, has expanded the teams capability to really get to the heart of data problems. This one is particularly tricky though, some ideas and leads but not as clearcut as I was hoping. Definitely one for the masses, and with a few clicks and a problem frame the issues is posted to the online competition board, accessible to millions of brainiacs around the world to pore over through the night. With a reasonable prize for the top 3 replies, you are confident that by morning there will be half a dozen brilliant responses to move the investigation further. Using the wisdom of the crowd has unleashed the productive capacity of the refinery.
Let me know what you think? I’d love your feedback. If you haven’t already then sign up for a weekly dose just like this.
Links that made me think…
Juicero is still the greatest example of Silicon Valley stupidity — CNET — www.cnet.com
Wild-eyed commentary: A year after a hyped-up, Wi-Fi connected juicer failed spectacularly, Silicon Valley’s obsession with it still makes me crazy. (Thanks to Cherilyn for this great link!)
Unmanned ‘ghost’ ships are coming — theconversation.com
It’s not all plain sailing when it comes to autonomous ships — they could make accidents at sea more severe and even end up being more expensive to run.
Real Work vs. Imaginary Work — Signal v. Noise — m.signalvnoise.com
Since we launched Hill Charts in Basecamp we’ve been fielding many interesting questions. One common question is: how do we catch more problems in the uphill phase so they don’t surprise us later…
Mystery of the cargo ships that sink when their cargo suddenly liquefies — theconversation.com
We know how to stop solid minerals converting to a liquid state mid voyage — so why does it still happen?
What to Do When Algorithms Rule — Behavioral Scientist — behavioralscientist.org
The first American astronauts were recruited from the ranks of test pilots, largely due to convenience. As Tom Wolfe describes in his incredible book The Right Stuff, radar operators might have been better suited to the passive observation required in the largely automated Mercury space capsules. But the test pilots were readily available, had the required security clearances, and could be ordered to report to duty.
