Today’s automobile development process is highly complicated. In the past, vehicles were just a combination of mechanical pieces. Now, automobiles are a multi-faceted combination of mechanical parts, electronics and in-vehicle software. As automotive designers, Honda R&D must continuously adapt our development processes to handle these complexities.

Many design problems actually occur early in the development process, but don’t become apparent until the latter stages. This results in many reworks at the end phase of automobile design. We need to improve the precision of the conceptual design phase to minimize revisions later. When design problems arise, we document them, of course. But often this documentation is in the form of text-based, unstructured data, which is hard to analyze systematically.

At Honda R&D, we proposed using AI to support automobile development by analyzing the unstructured data captured during the design phase.

Design thinking methodology drives to the heart of the problem

Rather than trying to apply AI to every problem simultaneously, we wanted to find out what our designers’ major pain points were, and how they hoped to change automotive development.

We invited our designers to an IBM Garage workshop to identify areas that we could address by creating clear scenarios and setting achievable goals.

In the workshop, we identified our “Countermeasure Request Form” as a sort of hill we could tackle effectively. This form documents a variety of issues mainly found by the manufacturing department during the prototype manufacturing phase.

When our designers select and check previous countermeasure request forms, they can see past manufacturing and design reworks. However, because these forms are managed in an unstructured data format, with descriptions of problems written by different individuals, they are difficult to analyze. Our designers often must seek out the individual authors to understand the problems described.

AI and IBM Watson technology – like asking an expert engineer

Our design thinking project evaluated the abilities of IBM Watson AI technology to analyze the unstructured data in the countermeasure request forms. I also built a machine learning model using IBM SPSS Modeler software to supplement the Watson capability. The project showed that our designers can set up appropriate tags to classify problems and search on them through a virtual assistant user interface. This search capability can help identify past issues to reduce some of the design reworks we currently encounter.

We used IBM Watson Discovery and IBM watsonx Assistant to analyze our text-based development documents and to create the virtual assistant prototype. We plan to apply the prototype to other areas, such as PLM, to improve our development activities for future models. This lifecycle of knowledge will be part of our weapon to further minimize reworks.

We also have a large amount of data in enterprise data stores, but we couldn’t search them effectively. The trial AI system shows that Watson solutions and AI models can recognize and classify our documents and give them appropriative tags.

Because AI learns from experience, the system will grow up like an expert engineer, and we can ask it for whatever information and knowledge we need. We’re basically getting an expert engineer in the form of an AI system.

The future of AI-assisted automotive design is here

I believe that the application of AI capabilities into automobile development has just begun. And I definitely think that it will apply not only to Honda R&D but also other manufacturers, other companies and other industries, too.

If a company is just starting to think about applying AI to the design process, I believe that having a design thinking workshop is the best way to make the real pain points clear and to create scenarios, capabilities and targets for the activity. A critical point is that the real users—engineers in the company or customers of the solution—should attend and contribute.

Right now, I’m one of only a few IT engineers at Honda R&D working on the AI project for automobile development. We are discovering the possibilities and the difficulties around using AI. It’s exciting to think that my work has the possibility to change the automobile development world. I find that thinking deeply about this issue is the greatest experience I can have as an engineer!

Learn more about how Honda R&D is transforming automotive design with IBM Watson AI technology by watching the video interview with Yoichiro Komatsu:

Watch the video

Was this article helpful?

More from Manufacturing

10 manufacturing trends that are changing the industry

5 min read - Manufacturing has undergone a major digital transformation in the last few years, with technological advancements, evolving consumer demands and the COVID-19 pandemic serving as major catalysts for change. To maintain their competitiveness and overcome today’s challenges, manufacturers have had to make agility and adaptability top priorities. Here, we’ll discuss the major manufacturing trends that will change the industry in the coming year. 1. Digitalization and Industry 4.0 Digitalization has had a profound impact on the manufacturing sector, enabling businesses to…

The future of order management solutions: freedom of choice and flexibility

5 min read - In the wake of the pandemic and global supply chain issues, businesses have realized the importance of technology innovation to deliver truly superior retail customer experiences. But without real-time reliable views of inventory, shipments and automated order orchestration processes, retailers are unable to deliver on order promises. Businesses need robust order management solutions (OMS) that can drive customer satisfaction, increase fulfillment profitability and support new digital and in-person customer experiences. These solutions must enable businesses to pivot quickly to support…

The missing link: Why visibility is essential to creating a resilient supply chain

5 min read - Supply chain visibility has been the missing link since the shockwaves of 2020 rippled throughout the world and consumers felt the impacts of broad-based supply chain issues. But what does supply chain visibility mean? It’s generally defined as the trackability of parts, components or products in transit from the manufacturer to their destination—with the goal being to improve and strengthen the supply chain by making data visible, actionable and readily available to all stakeholders, including the customer. While it’s clear…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters