Ensuring Quality in a War Zone

 Don’t miss the author’s follow-up to this article, “Going Low Profile in Afghanistan [1].”
--Editor

Profitable manufacturing requires efficiency, and this principle guides most (if not all) of our quality management systems (QMS). We’ve learned there’s little gain in correcting a production error if we don’t also find and uproot the error’s original cause. We’re so accustomed to this universal truth that we’re no longer even conscious of it. Sometimes, extraordinary circumstances can shock us out of this mindset and remind us that a quality system’s ultimate goal isn’t a more efficient organization. It is a quality end product.

ADVERTISEMENT [2]

For the better part of a year, I’ve been in southern Afghanistan, designing and implementing a quality program from the ground up. My background is in manufacturing, but war being what it is, I was pressed into service evaluating bridges, roads, culverts, and a number of other light infrastructure projects. Shifting into this civil engineering role was challenge enough, but the greatest difficulties were the cross-cultural friction and the realities of a war-zone environment. Oddly, one seems to mitigate the other, and the results in my case turned out to be instructive to manufacturing operations.

Ensuring quality in a war zone

Upon my arrival, we had no quality department whatsoever. The company I was contracted with had its own contract with a foreign-aid agency to build light infrastructure works throughout Afghanistan’s Kandahar province as a means of creating employment and improving the Afghans’ quality of life. Depending on how one wants to compile the statistics, Kandahar is the first- or second-most violent province in Afghanistan, along with neighboring Helmand province. The rural areas especially are where this type of aid is most difficult to implement, but also where it’s most needed. Our team had one year to prove we could create infrastructure within the chaos.

I sat down with a blank piece of paper to write our quality plan. I toyed with the idea of incorporating ISO 9001 principles into it, and I wondered if the old MIL STD 105E or MIL STD 414 plans could somehow be adapted to my situation. (Not really, as it turned out.) I’d been learning more about lean in the months leading up to this, although that didn’t seem to fit well, either. However, the 5 Whys seemed to hold some promise.

While still sketching out my plan, I met the Afghan engineers who would be my inspectors. These gentlemen were civil engineers, each with more than 30 years’ experience in their home country. They have forgotten more about concrete than I will ever know, yet somehow they didn’t know what an inspection report was. The concept of a notice to proceed (NTP) was also alien to them. The construction work was to be carried out by local construction companies that we hired as subcontractors. They, too, had little understanding of the aims and purpose of a quality department.

I could see that the ambitious, efficiency-producing elements of, say, an ISO 9001 system were going to be impossible to implement at this time. I was going to have my hands full just getting projects inspected in a timely manner. Meanwhile the Afghans could see no point in all this, and contractors wanted to charge ahead and complete projects faster than we could inspect them.

We were faced with a particularly ambitious contractor early in the program. We wanted a check dam built across a dry wash. This structure consists of a strategically located stone wall that can, during the brief rainy season, trap rainwater and recharge the local groundwater. This is water that would otherwise be lost to runoff. After a careful survey of the surrounding topology, a check dam’s height is calculated so that it will hold back a predetermined volume of water. Such a structure is simple, and the ambitious contractor had it half-finished before anyone even knew he’d started. Unfortunately, he built it in the wrong place. Given the slope of the ground where he put it and the elevation of the surrounding terrain, too much useful water would be lost.

Fortunately, the terms of our contract with him required that he rebuild it, and at his expense, but we didn’t want this sort of mistake happening on every project.

The author examines a flood protection wall along the Arghandab River.

Document control and the learning curve

Project management came to the rescue. We established schedules of progressive payments and coupled these to key inspection points that I identified for each project. Without a proper inspection and an NTP, a contractor would not get paid. This got the message across, whether the contractor appreciated the purpose of an NTP or not. It also established a minimum permissible number of inspections for each project. More inspections, as time allowed, were always welcome, but identifying the key inspection points for each project gave my inspectors a basic starting point to work from.

Although this system solved some problems, it also introduced others. I usually dislike linking the quality department to any kind of payment—we need to remain objective, after all—and our situation quickly demonstrated why. Our corporate office, not located in Afghanistan and communicating with us only via e-mail, somehow mistook the NTPs for payment certificates and could not understand why things like financial information weren’t being recorded. This misunderstanding was surprisingly difficult to clear up, and ultimately required a brief meeting face to face.

Aggravating the headquarters/field misunderstanding, our Afghan engineers were less than exacting with their paperwork. I was always on hand to catch their mistakes and advise corrections, but I couldn’t apply too heavy a hand. To succeed, our aid had to serve as a training opportunity, which requires paying attention and learning from mistakes.

This is how capacity building is done, and building the capabilities of the local people we work with continues to be a major goal of all the aid agencies operating in Afghanistan. Unfortunately, what’s good for capacity building can be a liability for documentation. As the inspectors filled out their reports and issued NTPs, I continued to coach them as to what they should report on and how their information should be presented. Their early efforts were clumsy, and it was important that they were allowed to remain so. Later reports were much more methodical and clear in their purpose.

By talking to the community and finding out what they really wanted, rather than dictating to them what aid they would receive, the structures we built will be used and appreciated for generations.

Root cause and the short term

By this point, my quality plan had shaped up nicely. I had key documents identified, a robust archiving system, methods for identifying inspection points, and even procedures for training new inspectors. The plan lacked only one thing: I was still wrestling with how to best incorporate root cause analysis. It’s not every day that one has the opportunity to build a quality department from nothing, and I wanted to get it right. But being as removed as I was from anything familiar, I just couldn’t anticipate what this analysis would look like in practice. I couldn’t figure out what in my manufacturing bag of tricks would be appropriate here. I decided we would simply have to wait for some nonconformances to turn up and then figure it out as we went along.

As construction on various projects progressed, and steadily-improving inspection reports rolled in, a surprising discovery was made. Between the inspection reports and the NTPs, we had everything we needed. Problems certainly came up from time to time—one contractor performed some sloppy masonry work that had to be completely redone—but searching out root causes of these problems often turned out to be a useless exercise. We needed to solve immediate problems, but analyzing their deeper causes was found to be purely academic.

This runs counter to everything we do in quality. The work of eliminating a root cause pays off only in the future. But our program had no future—at least, none to speak of. At the end of the year the contract would end, and we’d all go our separate ways. Our aid-agency client had no interest in seeing us streamline ourselves; what it wanted to see was functioning infrastructures. Also, we weren’t manufacturing 10,000 of the same widget. Our goal was to turn out a collection of unique check dams, bridges, and roads. No two jobs were alike. In the rare cases where there isn’t any future, quality end products are more valuable than efficient processes.

Goodwill and other valuable end products

The actual infrastructure items we built paid off in some valuable ways. Thanks in part to a culturally adaptable approach (which in itself could be its own story); our team was about the only one in the south that met with regular success. Although the security situation prevented others from getting projects started, we routinely got them finished, and did so in remote, rural areas thick with enemy activity. Word of this got out.

Word also got out that our stuff was built well; we sat on our subcontractors with inspection reports and NTPs, and made sure things were built as they were supposed to be. Before long, we had village elders turning up in our offices, asking us to rebuild the irrigation canals in their villages. These people were appearing from districts like Panjwaye and Rig; hostile places very far from Kandahar. In some cases they had travelled three days to see us, on the off chance we might be able to build something for their people. Ordinarily, elders like these would never speak to Westerners, but thanks to the quality of our output, they literally beat paths to our door.

In Afghanistan, as everywhere, customers, tribal elders in this case, care about the quality of the product you deliver. Your reputation precedes you.

The elders carry a moral authority that even the Taliban respect and they could probably end Afghanistan’s war tomorrow if they really wanted to. In a counterinsurgency, this is how we win hearts and minds. In manufacturing, it’s how we win customers. A customer doesn’t care what our processes are, except insofar as good processes turn out a good product. Good product valued by customers is ultimately what builds an organization. There is a point of diminishing returns to root cause analysis that we should not forget, lest the quality department become the least streamlined part of the organization.

Editors note: Next week we will feature another article by Alamanach on the right way and wrong way to handle customer relations in Afghanistan. The lesson is universal.

Be sure to leave your comments on this story by clicking the Comments link below (you may have to log in first).