Artificial Intelligence and Bots to Increase Engineering Productivity

Today, any conversation around innovation is incomplete without a due mention to ChatBots and AI. As per DARPA, we are entering the third wave of the AI which will be defined by contextual adaptation. We are witnessing the technology companies and startups putting huge amount of investments in this area.

A layman's definition of AI is still vague and people argue whether something is a ‘real’ AI or not. As John McCarthy, the father of AI famously said- as soon as it works, no one calls it AI anymore”. When the Deep Blue defeated Gary Kasparov for the first time in 1996, it was the rise of AI, but today we play chess everyday on our smartphone and see it as a computer application making use of brute-force computing power.

Continuing the same point, the interactive ChatBots that we see today- are they using real AI? Well, some are, and most aren’t. At a conceptual level, we can classify the ChatBots based on the type of response that they can generate and the breadth of context that they can handle. The ChatBots that we see today are covering this full spectrum- from being just a sophisticated way to answer FAQs to the ones which explore full AI tool-sets. Figure below shows the entire spectrum and what the current state of each type is.

As most of us would agree- Hybrid bots which use a combination of generative and retrieval based methods to respond to user queries are promising in the near future. Does it kill all your excitement about ChatBots? Well, it should not. Though the Bots that we see today in production are not truly generative and open domain, they still are a great help to improve the productivity and enhance the user experience of the end users.

While use cases that we see today are catering to the various areas of customer engagement, I think this technology has a great potential in product engineering. This can potentially help the product engineers to achieve their Utopian ambition of “truly concurrent product development”. 

In a hypothetical scenario as depicted in the figure below, a Virtual Orchestrator ChatBot liaises with engineers who are responsible for engine & HVAC design in a synchronous manner to finalize the HVAC design for a new automobile model that is being developed. The Bot then sends the final design across to the supplier asking for the detailed plans for prototyping and testing.

In the real world, the Virtual Orchestrator ChatBot will need to access different design/simulation tools and data sets which are required to perform the necessary actions. Though it looks a bit futuristic, with the rapid development that we are witnessing in Bots, AI, and data consolidation technologies, this is not too far from reality. At Capgemini, we are leveraging our BOT and Automation framework, strong engineering expertise, and our fail-fast Devise methodology from AIE to work on some Proof of Concepts in this area of Engineering Bots already!