Why Things Fail

The recent Boeing 737 crashes caused me to look up other product failures, a few of which I’ve worked on after-the-fact. Product failures are a regular occurrence, but I think many of us, myself included, are somewhat desensitized to them because we have so much going on and it happens so often. We make a mental note not to buy phones that blow up and move on.

There are thousands of decisions made during the lifecycle of a product. Many of these greatly impact time-to-market, productivity, cost, and operation of the product in the field. It is important to ensure that these decisions are the best they can be given the available data. The complicating factor is these decisions are frequently made by numerous stakeholders in many isolated domains, often in many parts of the world.

All organizations design products with the intention that they’ll be successful. Many industries spend a great deal of resources on testing, verification, and validation. Some even carefully document requirements and compliance to applicable regulations. They have deployed expensive systems such as PLM to manage their products throughout their lifecycle. No one intends for catastrophic failure, sometimes resulting in loss of life, financial losses, and substantial damage to their brand, but still it happens. However, things generally don’t fail by accident—regardless of intent to the contrary, it’s a choice to fail.

Here is one example from an organization I worked with, which I believe is more quality-conscious than many industrial companies: NASA. The Columbia space shuttle disaster was the result of foam that broke during takeoff, damaging the heat shields. A Columbia investigation board member, General Duane Deal, summarized it powerfully: “The foam did it … the institution allowed it.”

Failures are the consequences of mistakes at the decision-making level, where there are really three areas of importance: ideas, issues, and rationale at critical decision points. Failure is either the result of rushing into a bad idea, tackling an issue with a terrible practice, and/or perpetuating these at critical junctures in the product’s process. Once you push the first domino, you set the chain in motion. If you let it, it will heedlessly run its course right over the edge.

Bad ideas are often the first idea left unchallenged. This is the path of least resistance, where a decision-maker will accept the first idea due to convenience and/or pressure for rapid action. Potential issues may not be considered and the data won’t be consulted.

But once that idea is committed to, everyone is trapped in the vicious cycle of propping it up, like maintaining a wall of sand against the tide. This can go on for years and years. The longer this goes, the more they fall into the “Sunk-Cost Fallacy”—in which organizations rationalize their continued missteps to gain something—anything—from all their wasted efforts— Managers are often unwilling to accept the blame and will likely pin failures on external issues such as government regulations or unforeseen budget cuts. Subordinates will often not challenge them for the sake of loyalty or conformity. These evaluations are defensive and self-harming, and consequently, there is no push to divert time or money to investigate the original bad idea and its objectives. Failure becomes inevitable.  

If you ignore the data, you likely don’t have the analytics on your failures, and if you don’t measure your failing, you will keep failing—the definition of insanity. As crazy as this sounds, many companies are incredibly inwardly focused, far more intent on getting something out the door than understanding how it works or doesn’t work in the field. Making decisions based on flawed assumptions is a common form of failure-based decision-making.

Another common flaw is basing decisions on premature conclusions as opposed to data that provides more alternatives. Often, you’re missing key data, so you’re coming to premature conclusions on partial information. And then, regardless of the amount of data, if your bias takes over and you spend time and money on the wrong things, your products will fail.

Most bad decisions are directly due to a combination of the lack of available data and poor culture, but there’s also a lack of strategic thinking. Due to a lack of planning for tomorrow, many companies will focus on optimization today, leaving themselves unprepared for tomorrow.

Anyone can optimize today and fail to discover tomorrow

Many companies, plagued by bad decisions and product failures. have PLM solutions. They may be filling out Failure Mode Effect & Analysis (FMEA) reports and possess a Quality Management System, but my contention is that means that they can’t be using their PLM systems effectively or dealing with the actual data. If they were, they wouldn’t routinely make bad decisions.

Instead of knowledge and decisions—including process decisions—being in the minds of people in isolated domains, the knowledge needs to be accessible to everyone, regardless of domain, across the lifecycle and throughout the extended enterprise.

The advantage of an open, end-to-end digital product platform is your ability to democratize all product data, from concept to end-of-life, across all domains, and throughout a dynamically changing supply chain. This is the growing reservoir of your product intelligence. More people are using functions like Visual Collaboration across disciplines and organizations to capture and share intelligence regardless of file types or other impediments.

Think of the product as the sum of its decisions across a Digital Thread, which has traceability backward and forward. The biggest time-waster today is people trying to find data. A fundamental element to driving good decisions is making data available to the right people at the right time.

As an example, if your asset in the field has a real “configured” Digital Twin, you have near real-time data to make dramatically more informed decisions. It’s a lot like seeing traffic congestion on Google Maps or Waze; having near real-time data gives you more alternatives. With that said, the vast majority of people that think they have operational Digital Twins (87%) are only talking about streaming data from an IOT platform and lack a truly “configured” Digital Twin.

If you’re using data-driven decisions, especially from an open digital product platform, you can reduce isolated systems. This not only helps free up the data while building up your product-related IQ, it cuts your IT spending on sustaining systems and allows more of it to be applied toward innovation projects—a good iterative early form of failurethat lowers your risk of product failure in the field and produces product innovations.

It’s a choice to fail.  Be the organization that uses an open digital platform—one that makes good data-driven decisions, avoid the traffic, and discover your tomorrow.