Machine Translation for better Product Development

Dealing with foreign languages has gotten so much easier recently. Even free translation tools produce superb results. Machine translation is a great example of the digital transformation that we are experiencing.

But why should we stop at translating text into human languages? Languages for describing systems, like SysML, have been around for decades. The method of applying such languages is called MBSE (Model Based Systems Engineering). While MBSE can have benefits when developing complex products, it requires significant investment for tools, training, consulting and configuration.

Model Based Systems Engineering (MBSE) Without the Overhead

Semiant provides you with the benefits of MBSE, but without the investment. Semiant “translates” specifications into a machine-readable system model. From the model, it extracts useful information, makes suggestions or even performs mundane but important activities. This prevents waste, reduces risk and speeds up development.

Semiant automatically “translates” the specification into a machine-readable system model. It uses this model to help you make the specification even better and to derive as much value from the system model as possible.

The idea of automatic translation into machine-readable languages is not new. For instance, asking a voice assistant like Amazon’s Alexa would first translate the question or command from the user into a query that the system would then execute. The user never sees the (automatically translated) query, only the outcome. This is how Semiant works, except that it translates the text into a systems modeling language.

Alexa is a general purpose virtual assistant while Semiant is a product development specific virtual assistant. The following figure shows the similarities between Amazon’s Alexa and Semiant:

Tangible Benefits

This approach allows Semiant to immediately provide the benefits of MBSE to the practitioners without having to learn a new modeling language or tool. It provides value that is convincing to decision makers as well. Specifically, Semiant:

  • Prevents waste: For instance, creating a glossary (or controlled vocabulary) makes sense to support compliance activities. But a glossary created by a human will almost certainly be incomplete and outdated on day one. Semiant will create it automatically, and it will always be up to date.
  • Reduces risk: Having an understanding of open issues reduces the risk of slipping of delivery dates, organizational risk due to undetected gaps in compliance and so on. Semiant even makes suggestions on how to fix issues.
  • Speeds up product development: By automating mundane tasks, like the creation of the initial traceability, you get more done faster with your existing team. Onboarding efficiency will also increase, as new team members get answers to many early questions from Semiant, rather than jamming expert’s time.

More Skills on the Way

Today, Semiant can extract glossaries from specifications. The system model is the foundation for this capability. But in the same way that MBSE has many benefits, we’re just scratching the surface. We will soon add more and more Semiant “Skills” over time to unlock the full potential of MBSE, without anybody having to learn a modeling language.

Photo by Ivan Bandura on Unsplash


The Key to Compliance is Traceability – Would You Like to Automate it?

Traceability has many benefits in product development. It is also mandatory for regulatory compliance, no matter whether you build cars, trains, planes, medical devices, industrial tech. There are very few exceptions.

Please fill out our 30-second-survey below. Thanks!

Traceability in product development is everywhere:

  • From regulatory compliance documents to your system description
  • Between user requirements and system requirements
  • Between requirements of any kind and tests
  • Traceability between risks and functions
  • … and many more

A Glossary is a First Step

Semiant, our automated quality assistant, already provides you with traceability between glossary terms and requirements. This is already useful for compliance activities:

  • Regulatory compliance requires a controlled vocabulary and a common language, which Semiant helps you establish.
  • The Glossary provides traceability between glossary term and requirements relating to the term.
  • Last, it provides traceability between terms, e.g. generalizations or associations between terms.

Risks of Not Having an up to Date Traceability

For products that require functional safety compliance, not having a sound traceability can lead to serious problems:

  • Missing or incorrect traceability can lead to a rejected audit, leading to delays in product launch and expensive rework.
  • In case of accident that involve your product, there will be an investigation. Gaps in your traceability could lead to heavy fines, even prison for executives or bankruptcy for the organization.

As an example, consider the cost of the Boeing 737-MAX disaster: Hundreds of people died, thousands of planes grounded for over a year, loss of sales, and a heavily damaged reputation. The investigation uncovered neglect of fulfilling regulatory requirements.

Would you Like to Automate Compliance Activities Related to Traceability?

Traceability is useful beyond compliance. However, creating and maintaining traceability is tedious. That’s why we are exploring how Semiant could help you with this.

We prepared a one-question survey below. We’d love to hear whether any of these ideas resonates, or whether you have even better ideas.

Survey not showing? Open on separate page >>

Photo by JJ Ying on Unsplash