From SQL to NoSQL - Toyota Material Handling’s eight criteria for migrating to MongoDB Atlas

Profile picture for user Dominic Wellington By Dominic Wellington May 4, 2020
Summary:
MongoDB’s Dominic Wellington explains how lessons can be learned from Toyota Material Handling’s history for the technology decisions it makes today.

It all started 130 years ago with a hand-held loom.

You will know the Toyota name as a manufacturer of cars but it has a broader and richer history than just that.

A history that contains a useful lesson on innovation in technology.

In the late 1800s the founder of Toyota Industries Corporation, Sakichi Toyoda, then just 24, invented a new type of loom.

Unlike previous looms, the Toyoda wooden handloom required only one hand to operate.

It removed the unevenness of the woven fabric so it improved the quality. It also increased efficiency by more than 40 percent.

Image of a Toyoda’s wooden handloom
A Wooden Toyoda Handloom (Image sourced via https://global.toyota/en/downloadable-assets/corporate-archives/loom_wo…)

This was Toyoda's very first innovation, but not his last.

If he were alive today people might compare him to Steve Jobs, Bill Gates or Elon Musk.

After decades of excellence in manufacturing, in the 1930s the company he founded would even switch gears into automobiles. Eventually creating a separate company that would become the Toyota Motor Corporation.

Whatever Toyoda did it always had that same spirit of his first invention.

He strove for perfection not only in the quality of the end product but also excellence in the process that built it.

Toyota Motor Corporation itself became famous for the Toyota Production System. The manufacturing philosophy that aimed to eliminate waste and achieve ultimate efficiency.

The Toyota name is now renowned for the quality of its products and the quality of its manufacturing.

Toyoda's innovation has also influenced our industry: technology.

Historically, we'll look back at this era as being the equivalent as the early 1900s in software development.

We're still working out where all the waste is. We're still working out where the big opportunities are and what the right principles should be.

And in this new era, Toyota Industries Corporation is innovating again.

Applying the same obsession for great products and great processes to software.

The company that started out with Toyoda's loom innovation now wants to become the leader in Industry 4.0 (the term du jour describing modern logistics and smart, autonomous and safe factories in the future).

Toyota Material Handling Europe is building the software to help Toyota reach that goal.

Building something great requires a willingness to rethink how things are being done.

That's what Kiichiro Toyoda's team did in the 1930s, they didn't accept the current standard of waste or quality in the factories.

A new paradigm required a new set of principles. A new set of criteria.

If you're building a massive, complex and flexible internet of things platform, what's the most foundational piece of that software stack? The infrastructure.

Image of Filip Dadgar, Principal System Architect at Toyota Material Handling Logistics Solutions
Filip Dadgar, Principal System Architect at Toyota Material Handling Logistics Solutions (Image sourced via MongoDB)

The database infrastructure is like the machinery and factory design from eras gone past.

If a company wants to build something quickly at scale that is also safe and of good quality, the limiting or enabling factor will be what kind of infrastructure they built on top of.

Over the last couple of years Toyota Material Handling Europe's team had realised moving to the cloud was a necessity to enable that speed and flexibility. They were making the move with increasing urgency.

However, they weren't able to get all the gains they needed with their current SQL database.

The rows and columns were predictable and well known to the team, but they were also restrictive and brittle.

One of the big benefits of a move to the cloud is greater flexibility.

The relational data model was being left behind as other parts of the infrastructure became more flexible and responsive to changes in demand.

These trends risked turning the database into a bottleneck, a speed bump on the road to digital transformation.

They needed a new set of principles. A new set of criteria.

Toyota Material Handling Europe took the decision to move away from SQL and the relational database model and onto MongoDB Atlas, the fully managed cloud database service.

MongoDB is a powerful document-oriented database and known as the leader in what was once called 'NoSQL' databases.

'But why did they make that decision?' The dear reader bellows into the screen.

Wouldn't it be helpful if there was a four minute overview by one of the men at the heart of the project?

Well, by luck and magic here's a short video from Filip Dadgar, Principal System Architect at Toyota Material Handling Logistics Solutions, explaining the eight criteria his team used when making this decision:

And if, like me, you're more of a 1930s type and prefer your principles written down. Here are the seven criteria Toyota Material Handling used to makes it database decision:

  1. Performance. The thing has to fly. Good latency, scalable reads and writes to meet service level agreements (SLAs) and accommodate for massive growth in the future.
  2. Automatic scalability. Operations, and maintenance that enable developers to focus on applications, rather than backend database administration, even if the load on the database increases dramatically.
  3. Security and compliance. To safeguard highly sensitive business and personal data the infrastructure must have best in class security and data control features.
  4. Data locality. For Toyota Material Handling Europe, they needed to host in Northwest Europe or West Europe regions of Microsoft Azure, therefore, ensuring close geographic proximity to factories for latency-sensitive operations.
  5. Automatic backups and restore. To ensure customers have fully fully-managed data protection and disaster recovery, the database had to offer seamless backups and restore.
  6. Cloud agnostic. Should it be needed, customers should have the freedom to run anywhere, on any platform.
  7. Developer friendly. A flexible data model that helps developers focus on building new products, integrating new application functionality and staying ahead of market demands.
  8. Ecosystem for developer productivity. Widely available skills, proven best practices and access to great online learning resources for quickly upskilling new developers where required.

If you're still interested in finding out more about Toyota Material Handling Europe's journey and how the team use MongoDB Atlas on Microsoft Azure, you can watch the full 40-minute presentation here: