"Gang bang" as a management strategy

"Gang banging" in the context of management strategies is to throw every available resource at a problem.  The intention is to "get it done" and crossed off the list.

The first time I heard the term fall out of a manager's mouth  I thought, "You have got to be zhitting me!"  Machismo, testosterone, overtones of sexual violence and stupidity....all in two little words.

Why is it stupid?

The laws of diminishing returns suggests that beyond some point more people reduce effectiveness.  They start getting in each others way.  Lack of continuity causes mistakes.

Defacto evidence of incompetence

The target of the vast majority of "gang bang" efforts are entirely predictable, routine maintenance that fell into arrears.  Whether the maintenance fell into arrears due to poor planning, resources being sponged up by other tasks or a mindset that prevents the exploitation of small time windows is irrelevant.  Management failed to get the job done.

An example

I like trees.  I like trees in cities.  There are smart ways to plant them and there are dumb ways to do it.

The dumb way
The city gets a grant to replace all of the trees that died in the last ten years.  A salesperson makes them a deal on 300 Honey Locust trees.  The municipality "Gang bangs" the problem and drills them in.  There is one-chance-in-three that most of the trees don't make it through their first summer due to an entirely predictable (in Michigan) six week dry spell.  It is pretty easy to keep hoping that it will rain next week when you don't have the resources to water 300 trees.

The smart way
The city consults with local experts and develops a list of ten recommended species (or varieties) of trees.  Since the average street tree only lives 30 years one needs to plan on replanting about 3% per year.  In Eaton Rapids that would be about 35 trees a year.

Each year, 35 trees of a single species would be purchased and planted where that species is most likely to thrive.  The next year a different species would be chosen, such that over the ten year cycle the list would be depleted.

Celtis "Magnifica"

Species that grow large can be planted where houses are widely spaced.  Narrow tree species can be planted where the houses are closer together.  Drought tolerant trees can be planted on the sandiest sites.  Flood tolerant trees can be planted where standing water is likely, and so on.

While the City of Eaton Rapids would be overwhelmed trying to water 300 trees through a drought, 35 trees is pretty manageable, especially if accommodations were made ahead of time.

An example of a 33 year plan.  No single species is more than 3% of the planting and no genus is more than 18% of the planting.

If you have a good working relationship with the tree nursery, you can share you planting plans out five or ten years in advance, so rather than getting stuck with the dogs the nursery could not sell the city gets prime trees.

Another example, this time from industry

Welding robots are "dressed" with water hoses to cool the equipment and air hoses to operate clamps and (sometimes) weld guns.  Over time, the flexing causes the rubber hoses to crack and leak.  During production, the hoses are repaired with splices.  Each splice restrict the water flow and compromises the hose's ability to flex.  Consequently, the hose becomes more vulnerable to further cracking and leaking.

I have seen robots with hoses that had as many as five splices in a single hose.

The dumb way
Wait for an extended downtime and Gang bang it.  Get four pipefitters per shift out there.  Have them strip the hoses off forty robots and completely rehose them.  The reason this is dumb is because it is not obvious which nipple on the supply (usually near the base of the robot) corresponds to which nipple on the business end of the device.  Furthermore, a hose can be too short...such that it strains when going through its path, or a hose can be too long and floppy.

The smart way
Develop a plan and practice it until two pipefitters can rehose a robot in 10 minutes such that one or two robots can be rehosed during any lunch break.

"Impossible!" you say.  Picture in your head that the pipefitters pre-cut the hoses to 14 feet.  That is the length of the longest hose I have ever seen in the application.  They do one hose processing.  Disconnect one end, insert the end of a new hose of the appropriate color.  Walk the hose to the other end following the hose that is being replaced and cut the new hose so the length matches.  Remove the old hose from the second end and plug in the new hose.  Mark along the length where any supports or binding is present.  Repeat.

A robot might have as many as six hoses.  By replacing them one-at-a-time the pipefitter does not lose the information that is embedded in the original installation.

Hoses with excessive numbers of splices might benefit from a touch more length than the original...so give it six more inches.

One slick alternative is to not even cut the new hose to length but to coil the extra length, usually at the base where it does not move, and loosely bind the coils with cable ties.  Then the pipe fitter has the option of teasing out a few extra inches if the validation shows the hose being excessively stretched.

Validation...oh, right.  A team of pipefitters can rehose two robots in twenty minutes and use the remaining ten minutes to validate the work.  The two principal players are still next to the equipment and there is no dilution of responsibility.

Validation of Gang-bang work devolves into a goat festival with finger pointing, recriminations and down time.

Gang banging===> Stupid.