While I have spent my entire career in a post-occupancy setting, coordinating CAD/BIM and technology projects on an Engineering team for a large medical facility is much different than administering a CAFM/CMMS for an equally large (though much more far-flung) office environment.
This is my final installment in that series, tasks which I have begun working on, and which I will continue to plow through. I will link to the other articles at the bottom of this post.
I won't bore you with any more talk of goal-setting after my last article. But, of course, I managed to tie my personal goal up into one of the goals for my department. We aim to improve our stewardship of the assets in our care this year, by changing the way that we track and monitor them (in some cases that means collecting more data, in others, it just means taking time to analyze that data to see what bubbles to the top and can inform our choices. Applied analytics or organizational informatics is what this is... oh, yeah, big data, baby!).
My personal aspiration is to audit every SQL data table in my Archibus installation.
585 tables. I've audited 26 so far.
That is a bit misleading, though. I might open a table to see what it contains, and detect where the data gaps are. But, a true audit will only come about when I reference tables against each other. That is a little harder to track, but, I have a future post in mind about how to streamline this for long-term use.
Examples (skip this if you're not a hopeless nerd who gets off on organizing and trending massive datasets, seriously, I won't be offended)
Our real estate manager routinely runs reports on our active leases. I can't even conceive of all of the metrics she has established for these, but, one of the more important data points when querying for her report is whether a lease is active or inactive. Her regular audits and established processes ensure this data is always kept accurate.
However, each of these leases is also assigned to a building. I found by looking at the building table that no one was currently marking when a building was no longer being used. This was evidently done by another employee in the past. This can have repercussions in all of our modules. So, I created a process to address this, and had myself inserted into the move notifications, so, the project manager who moves us from one location to another sends out a mass email to those who need to know, which now includes me. At the appointed time, I mark the building inactive and move the Work Requestor employees from the old building to the new one, ensuring maintenance and other records are applied to the correct location.
I mentioned in Part 3 of this series, that I detected some problems in our maintenance KPIs (key performance indicators), in that there were too many generic work requests, which could have fit into existing categories for more granular and accurate tracking. Making things easier for the (non-maintenance) work requestors involved paring the main list of problem types from 32 down to 16, and renaming some to be more clear (such as switching 'Mechanical' to 'Heating & Cooling'), and adding sub-types for common issues.
As we were having this discussion, we actually talked about having keys made, which we did outsource, but, after I asked our corporate facilities manager how much we spent on that, she got a quote for a key machine and we saw the ROI was at 18 months. Bam, now one of the guys makes keys in-house for us, saving both money and time (that's like $^2, right?).
That is analytics informing business decisions and process change. Beauty.
I recently streamlined the room types table, but, it could probably still use a bit more tightening up, I need to consult with our corporate facility manager and vp before I make any further changes, as they are the experts on how our buildings are currently used and might be in the future.
I mentioned above how I will only have a complete audit when I reference some of the tables against one another. For example, I can look at the rooms database and see if there is a space number which isn't linked to a dwg (an orphaned space designator after a renovation, or a number manually entered during an employee move which doesn't actually exist?), and I can look at the employee table and ensure all information is filled in. But, it isn't until I pull these two tables together that I see that maybe there are three employees assigned to one cube, and I know we're not THAT crowded. That is the point where we chase down possible causes and implement new procedures, or tweak the views where users input data.
Rather than going on and on with similar examples and possibilities, I'll wrap up now.
Have you done some data auditing? Did it help you prevent mistakes? Save you money? Cause you to outsource or bring new functions in-house?
Upgrading Archibus from release 18.1 to 21.1, and after some unresolved errors, ultimately to version 21.2. Migrating data, forms and users, documenting as I learned about our processes.
Learning how our real estate team works and customizing a form with numerous panels, in order to keep the information together and easily accessible... and the crazy unexpected and hard to reproduce errors that slowed us down for a few weeks.
It's important to build good rapport with your users and remind them to report errors as soon as they happen (good users try to work around problems, but, that doesn't solve them).
Adding 500 branches to our CMMS (for work orders, we'll tackle preventative maintenance later on, the branch team is still streamlining our hvac contracts and working up standard procedures for our equipment) and creating a simple end user tutorial on entering work requests.
The whole branch team now has an SLA and is using the new Operations Console in WebCentral, but, I still need to migrate our corporate team from the old pages, their user roles and VPA restrictions will need to be wiped and started over, as there is a lot of customization on their current settings which prevent their seeing content in the new Console.
Archibus's bread and butter is space asset management. Migrating tasks and functions from the thick client to WebCentral and introducing the Space Console and making plans to capture common spaces not currently polylined.