A robust and reliable product development process can make all the difference when developing new products in an efficient way. Oftentimes, however, groups may find themselves unsure of how to push a product through to development as a result of constraints on time, money, and equipment. This is usually the problem with pipelines that do not take full advantage of new, holistic approaches to development that offer cost-effective solutions to these issues. At Windpact, we believe that adaptive approaches are best used to address these pitfalls, and our use of data-driven product design has proven time and time again that holistic perspectives are powerful.

Without the right setup, implementing a new product development process can itself be a stressful undertaking. This setup includes gathering data on the breadth of materials to be worked with, and upfront investments in predictive modeling infrastructure that will offset the costs of prototyping down the road. Though the initial costs of this preliminary work can be discouraging, we believe the costs and time saved at later points of the development pipeline will create massive returns.

As part of our belief in the power of data-informed and holistic modeling approaches, we will explain what this new process will look like, broadly named Industry 4.0 for the Fourth Industrial Revolution.

Material Data Gathering

Collecting data on the types of materials you work with is one of the most pivotal steps towards running an efficient product development cycle. Otherwise, your pipeline lacks the informed approach that is based on real, measurable metrics. Thankfully, we live during an era where data can be collected from nearly anywhere, and massive databases are commonplace for groups that need to handle large amounts of data points. For example, technologies taking advantage of the Internet of Things have immediate access to data points that number in the thousands, some even in the millions. So, clearly the ability to derive data is not a pertinent problem. The tricky part is getting access to high-quality, pedigree, data, and testing the integrity of your data sources; if you draw data from anywhere, without verifying its quality and integrity, you can run into even worse problems.

Gathering high-quality data is critical both from a marketing and engineering standpoint, and you will always be gathering data before and even after product launch. This includes preliminary data such as material specs, analytics, performance analyses, among many other technical data points that matter for your product’s individual creation and optimization. Beyond that, you’re also working with data derived from the market itself: where in the grand scheme of your industry does your product satisfy a specific consumer need, or more specifically, in which direction is your field heading in terms of technological innovation?

Every industry faces challenges in verifying data, and the consequences of bad data collection are well-known. In our last blog post, we analyzed the ramifications of Trek Bicycle Corporation’s bad publicity regarding their exaggerated product conclusions. This is one of the more recent examples, but they’re more common than you think. Product design in the safety equipment industry itself presents unique challenges.

For instance, in materials science, engineers need to take into account the basic physical and chemical properties of a material intended for use in a piece of equipment. Some phenomena, in this case, viscoelasticity, require significant testing in order to ensure that things like padding and synthetic polymers deform and reconstitute themselves in a manner that maximizes safety against physical forces. Viscoelasticity, briefly, is a property of materials in which a material reacts differently depending on the rate of deformation (“viscous”) , and then “spring-back” to their previous constitution (“elastic”). Many types of conditions can affect the viscoelastic response of a material, including force, load, shear, and temperature (for example, metals at high temperatures exhibit viscoelastic properties). Testing includes paradigms like the creep-recovery test, where materials are subjected to a constant stress for some period of time, then removing that load and measuring the time-dependent reconstitution of the material back to its original formation.

Now, a short story would be that everyone conducted these objective tests and that the resulting products’ ability to perform is proven. However, outside of just looking at the objective testing part, we find it helpful to see where that data was derived from aka its pedigree. Data pedigree refers to not only the history of the data itself (what and when was it collected), but also, the quality of the data (how accurate and representative is the data, truly). The pedigree of data, especially if you were not the one who directly collected and cleaned it, is important information to reference when drawing conclusions.

This is precisely how Windpact rises above the competition. Our novel process, and trade secret, is how we test a wide range of strain rates in viscoelastic materials and digitize the results. These materials include open cell and closed cell foam materials as well as additive materials. The manufacturers of these materials have a general understanding of how their materials react on the low and high end of the strain rate spectrum, however, most have little to no knowledge of how these materials behave in the middle of the spectrum, the area most critical to blunt impacts. The resulting pedigree data is our core advantage to solving complex design and impact problems.

Predictive Engineering

Let’s say that you have gathered data on your materials, and that the pedigree is sound; the next step is to apply those data towards further testing, and ultimately, the creation and optimization of the product. We find that in the material engineering space, designers have a vast breadth of tools at their disposal to create prototypes and generate products all the way to their launch. Our team at Windpact, therefore, believes that tools that emphasize predictive engineering are some of the most powerful.

Predictive engineering, broadly, is the preemptive simulation of a product design’s performance, deriving data from computational models so that the framework functions as a proactive solution, rather than reactive updates. Refining the simulation based on testing protocols already in place with predictive engineering principles allows for better reporting on data analytics, as well as setting you up with a framework that is amenable to changes over a product’s lifetime.

Predictive engineering answers the need for engineers to continuously update their product design, and responds to the constant desire for teams to consider new materials and configurations based on market climate. Having abundant materials data at your disposal, combined with the cost-effective power of predictive techniques allows for faster time-to-market with fewer roadblock. There are different types of engineering tools that fall under the predictive engineering umbrella, including:

  • Finite element analysis (FEA): process by which computational models simulate a design’s predicted mechanical behavior under a variety of conditions.
  • Computational fluid dynamics (CFD): simulation of fluid behavior, important for measuring hydrodynamics and aerodynamics.
  • Deep neural networks: using data to train an algorithm that generates predictions based on its trained data set, important for measuring mechanical properties of composites.
  • Topology optimization: a mathematical method that optimizes material layout and geometry within a given design space, for a given set of loads, boundary conditions and constraints with the goal of maximizing the performance of the system.
  • Artificial intelligence (AI): computational method in which machines have cognitive-like functions that allow them to learn and adapt to solve problems

An effective product development pipeline will use some combination of any predictive engineering tool, and the needs of the team need to be considered when deciding which kind of process to invest in. Common challenges toward adopting a pipeline rooted in predictive engineering range from the high initial costs associated with developing a FEA model, training personnel, software licensing, and technical support, as well as the pressure to update when new techniques become commonplace.

However, the benefits of a more holistic approach far outweigh these challenges, and groups will soon find the costs saved on rapid prototyping and product optimization more than make up for the initial investments. Furthermore, our team at Windpact knows exactly how to take full advantage of a predictive pipeline, such that continuous testing and optimization are made straightforward and as easy to adopt as possible.

New Product Design Process Improvement Opportunity

To understand how far we believe a holistic approach to product development will take you, we need to revisit the current state of affairs. The current product development process that was, up until recently, standard, was straightforward at first glance. It involved:

The Idea: where all great products first start, the idea is the foundation around which you build your product. The way ideas are formed are dependent on the market you belong to, and the need for your product in your sphere. Some teams have various standards by which they pitch and greenlight initial ideas, but the end result will always be a decision that serves as a guide that directs how you move forward to a final, fully fleshed out product.

CAD & Design: Now, you make your design a reality by creating a model using computer-aided design (CAD) tools. Even at this stage, however, the problems in the traditional approach are apparent. Sometimes an idea, while it works well in theory, is simply not practical or realistic after its CAD is set. Back to the drawing board until a workable design is achieved. Rinse and repeat.

Prototypes: The CAD checks out, and now prototypes are being generated for testing. This requires tooling that once fabricated, has little to no ability to be modified, requiring additional capital investment each iteration. Again, you run into this problem of constant creation and revision, where you often revisit the CAD and idea stages many times over.

Evaluations: Even after a prototype passes quality checks, how they perform against existing tech as well as how they function when new materials are introduced to the market in the middle period between idea and prototype can force you back into the cycle of revising and testing.

Solution: Eventually, after multiple trials and errors, you get a solution that checks all the boxes (or at least most of the boxes as priorities are always weighted) for the goals of the product. This does not mean, however, that the product is optimized.

This process does not seem as straightforward now, and this is why we champion the use of a holistic approach. Even though you arrived at your solution, all of that backtracking to different earlier periods of the pipeline costs time and money. This also does not come with a guarantee that your product is truly optimized; all of that revision may have forced you to sacrifice certain aspects of the design or performance to arrive at your solution. Instead of compounding unexpected costs as they come up, it is far more parsimonious (and let’s face it, cheaper) to invest in a product development process that uses new approaches through predictive engineering.

This pipeline instead emphasizes:

The Idea: the same starting point for all great products, but hopefully you have some data or at least a general sense of what the market demand is for your product and how its intended materials will reflect a new, novel frontier on the industry.

Automated design and engineering: The critical point where we deviate from the previous status quo, and instead, rely on an automated process that includes predictive analytics and engineering.

Optimized solutions: because you used an automated pipeline that draws on data and takes full advantage of its design flexibility, you can arrive at a solution that is truly optimized.

Industries that could use new development methods

So why do industries continue to use old product development methods despite the clear advantages of new advanced approaches?  For the most part it’s due to a lack of funding priorities and understanding that to adapt to an agile mindset, you must first invest into the right tools.

It’s not because these groups do not value optimized solutions. Much of the problem with things like product pipelines is a heavy reliance on legacy hardware/software that simply doesn’t translate well into new technologies. If a key aspect of a team’s pipeline includes legacy technology, then the benefits of transitioning to an entirely new model for development may be hard to see in the long term.  Here are just a few examples.

Sports (protective hard goods category):  The sports hard goods industry, at large, is lagging behind Industry 4.0.   Most protective equipment brands are stuck in the, ‘this is the way we’ve always done it” mentality, which is why innovation has been incremental, at best. Let’s focus on the largest helmet market, cycling helmets. These helmets are prone to failing after multiple impacts, a result of the type of EPS material many of them use. This means that after only a few uses, helmets used for cycling have reduced effectiveness over time. Multi-impact scenarios could be tested and accounted for through more automated approaches in a cycling helmet’s design. The entire sports industry is starving for a breakthrough in technology and investing into CAE is the only way they will make a giant leap forward in innovation.

Military: Windpact and attest that the United States DoD is at the forefront of investing heavily into Industry 4.0; if they didn’t, we wouldn’t have the strongest and best equipped military in the world. As illustrated by Windpact’s contract awards by the U.S. Army Natick labs, we developed a revolutionary FEA model for the ECH helmet with industry-changing results.

Construction Safety Equipment: construction helmets, often called “hard hats,” are durable shells that have a unique suspension system, which spreads out the weight and force of impacts around the entire head. While this is great for preventing penetration through the helmet and ventilation, it is not the best configuration to reduce the severity of head impacts. A holistic approach could find areas of improvement for this suspension model, allowing for more targeted prototyping.

Automotive: The automotive, and aerospace industries, have long been the leaders in CAE and lean manufacturing technologies for decades.  After all, it’s not practical to continuously crash a Cadillac Escalade to gather data on how the vehicle will perform in a crash.  These industries understand everything there is to know about metals (ie steel, aluminum) and utilize the most sophisticated FEA models.  However, they do have a void with viscoelastic materials data and Windpact’s materials data library can close this blind spot in their product development toolbox.

Footwear: Given the explosive trend towards athleisure and performance footwear, how can a sneaker be optimized for a 200 lb male vs a 125 lb female?  How does this optimization change for a casual walker to an elite athlete?  Currently, the midsoles and foot liners in sneakers are a “one size fits all” with standard EVA materials.  We believe this is a category ripe for innovation with the right investment into Industry 4.0.