Myths Vs. Reality About Engineering Software

Over the 2.5 years that I've been building a business that develops simulation software for engineers, I noticed a couple of strong myths which dominate at engineering companies, and frequently prevent them from adopting efficient digital solutions. And since the business I’m leading chose a startup funding model, I have had a chance to dive into tech investors’ minds, where I also spotted a couple of myths, however, different. Now, I’d like to present a short summary of those frequently encountered myths and how I find the reality of the engineering software market instead.

Myth #1: Engineering based on experience is better than any software

This is what I’m constantly hearing from small and mid-sized companies in traditional industries like metalworking, machinery, metallurgy etc. And actually it’s true - a good engineer with 30+ years of experience is a much better “asset” than any software. But the point is - it takes 30+ years to develop that “gut feeling”. And what if you want to scale a business and need to hire more engineers? Or your “golden asset” leaves?

About 70% of all knowledge in the industry is stored nowhere else except in the brains of experienced employees, and about 50% of them are set to retire in the next 5 years [1]

Modern businesses just can't afford to wait 30 years to train a new employee. So, a sustainable business can not rely on the experience of one or a few engineers. Thus, software helps.

Myth #2: Engineering software is a luxury (expensive), only suitable for big enterprises

This opinion is very strong, preventing SMBs even from thinking about purchasing software – they don’t even dare look for it. There are two misconceptions behind this myth: 1) high price, 2) special engineering skills required to use the software.

1) The biggest value of engineering software is a library of smart algorithms developed by talented (and expensive) mathematicians. That is why such software costs a lot and has to focus on enterprise customers segment.

Community-based open source software has already been proven to outperform single-company-developed proprietary tools and conquered about 80% of Internet and software technology in the 21st century

Just think about Linux. That is the strength of the sharing and collaboration culture of smart academic minds, unleashed by the concept of open source. The downside of open-source is a limited user-friendliness due to its non-commercial purpose. Nevertheless, the truth is that there is no excuse for the high price of algorithms-based software anymore. One can get even better algorithms from open source and just charge for custom user functionality.

2) Because of the enterprise focus, typically, engineering software tries to be multi-purpose one-stop software for any kind of engineering applications, to cover a wider market. As a result, a specially-educated so-called simulation engineer is needed to set a case for a particular problem. As any narrow specialist, such an engineer is expensive and likely will not be able or willing to perform other engineers’ jobs. Nevertheless, the truth is that open source communities can easily bake focused apps based on the basis of common algorithms, one just needs to put together those who code and those who use, and provide a uniform framework for the apps.

Myth #3: Cloud Vs. Desktop

While cloud technologies revolutionized many markets by freeing end-users from hardware maintenance headaches, there is still huge resistance in the engineering world. Especially SMBs, who are contractors or suppliers (and therefore are bound by NDAs) are afraid to share any piece of customers’ information, be that a CAD file or any other operational or design data.

Nevertheless, there is a strong opinion among tech investors that everything will be moved to cloud. It is because 1) cloud solutions offer subscription and pay-as-go models, which eliminate up-front investments for customers (but it could be also easily implemented to desktop solutions, all you need is an internet connection); 2) a software vendor gets access to all usage statistics which is awesome to have to improve the service; 3) hardware is easily scalable in cloud to ensure enough power for any customer's task.

However, in reality, in the B2B sector, customers still prefer desktop solutions and this preference is not changing as fast as it was predicted. It has a real chance to go regress (or slow down significantly) because of the recent trend in data privacy and data policy issue of cloud service providers such as AWS [2]. Even more, since a customer starts to use software frequently it becomes cheaper to purchase their own hardware rather than pay for HPC services. But the most important bloody truth is that cloud is just one more obstacle in the mindset transformation of the industry toward digital solutions, so why do we need to insist on cloud if a customer says “no” to it? It seems like it’s because software vendors want to make their lives easier by avoiding installation issues and getting user data. But instead of that, we need to focus on things that really matter - quality of customer service and value the software delivers, keeping the form of least resistance to uptake. If it’s desktop - let’s give desktop!

Myth #4: AI will take over everything

I guess that everybody loves when Gmail helps us to finish a sentence. Am I right? No doubt, machine learning is a super powerful technology. But it just has natural limitations. You need massive data to teach a machine to make (even suggest) right decisions of design parameters for engineers. In many cases industry segments are just not big enough to get sufficient amount of data even if all design decisions of all engineers in the world are collected. Not even mentioning the fact that the majority of engineers don’t like to share their decisions.

While an engineer makes decisions not just based on experience and statistics, but mental calculations based their education in physics. Therefore, an engineer needs less data than AI to come up with the best design. Therefore, it is just an exaggeration to believe that AI will outperform specific design jobs where the data amount is not so big. Actually, in many cases, engineers need tools to be more efficient, not to be replaced.

Because of those myths and what we found the reality is, my co-founders and I launched CENOS - a desktop engineering simulation software, which integrates the best of open source algorithms to provide seamless user experience for engineers who need a tool to replace time-consuming physical prototyping and testing with a digital solution.

In the end, I just want to say:

• Engineers, do not underestimate software, it’s designed to help you to be efficient! Don’t be too busy for trying things which are here to save your time...
• Business decision-makers, do not underestimate software, it’s your chance to be more competitive. Perhaps, your last chance…
• Tech investors, do not overestimate cloud technology and AI. There are reasons, objective reasons, why some significant segments of the market don’t want such tech. Don’t fight them, rather don’t ignore them - it’s an opportunity!

[1] A.Barnwal (a global product marketing manager at Honeywell International Inc.) -"Transform your thermal process performance with digitalization" - Thermprocess Symposium, June 2019, Dusseldorf, Germany

[2] Press conference by Margrethe Vestager, Member of the EC, on Luxembourg McDonalds' State Aid case, 19th of September 2018, Brussels - EC/Berlaymont