Interview: Making EMF Intelligent with AI

Niranjan Babu's talk Making EMF Intelligent with AI was chosen as an early bird selection. Read this brief Q&A with Niranjan to find out more about the I-EMF project.

Q: How did you begin with the idea of combining AI with EMF?

A: EMF is used extensively as a modeling framework in the automotive world and the entire world is moving towards model driven development. These modeling frameworks are thus critical in determining the efficiency of software development. I thought the best way to improve efficiency is to make these models intelligent. That is when machine learning came into picture and I decided to combine machine learning with EMF.

Q: How did you begin with the idea of combining AI with EMF?

A: EMF is used extensively as a modeling framework in the automotive world and the entire world is moving towards model driven development. These modeling frameworks are thus critical in determining the efficiency of software development.
I thought the best way to improve the efficiency is to make these models intelligent and that is when machine learning came into picture and I decided to combine machine learning with EMF.

Q: How did you find the synergies between the AI domain experts and the modelling experts?

A: I have personal experience in developing Model based tools for over 10 years. In the past few years I have also had an opportunity to work on Internet of Things based system. In these systems we collected measurement data from cars at my organization. I realized that this collected measurement data could be used as learning data. Taking this knowledge that we might base learning on the collected data, we got the help of in-house AI and ML experts and we started working on these topics.

Q: What was your process for developing I-EMF within Bosch?

A: I-EMF is in its initial stages and we have built its first version. We are now promoting it internally at Bosch so that various business units use it. Based on the feedback and the different use cases that we get to work on, we intend to improve the features in the framework, so that all these use cases are supported in a generic fashion.

Q: Can you tell us about your users or perhaps their use cases?

A: I-EMF is used in ECU Software development tooling departments. Here we have a lot of data to learn from and this data is used to dynamically customize the tooling based on the work areas of the respective Business units.

Q: What do you see in 18 months for I-EMF?

A: I see I-EMF supporting multiple areas and several development use case such as design, simulation, autonomous driving etc. The next stage for I-EMF is to host it as services which can be used in distributed systems.

Our Sponsors

For information about becoming a sponsor, please visit the EclipseCon France2018 sponsor prospectus page.

Premium

Basic

Regional Supporter

Conference App Provider

Media

JUG Partners

EclipseCon Support Other Events

Our Other Events

Eclipse events are hosted all over the world!

  • EclipseCon Europe 2018