Mr. Carlos Agudelo, Director Applications Engineering at Link Group and Shravan Adapa, Director of Data Science, Soothsayer Analytics will be delivering a keynote address as part of the opening plenary session taking place on 17th May at 09:15 - 10:15am CEST. They will be talking on the highly topic of "Can AI help reduce testing from the test engineering process? How data analytics can make testing smarter"
What do you find most interesting about the topic you are presenting on?
AI and the associated technologies and methods have the potential to reinvent the field of system development, allowing the engineer to ask questions that current testing and development models cannot address
How has AI changed in the past 5 years? What do you predict will happen in the next 5 to 10 years?
We have come a long way since the workshop in late 1950 at Dartmouth College, New Hampshire to develop machines that could “use language”.
The last five years have seen a quantum leap in computing power, integration of disciplines, and technology systems and platforms (from IBM’s Watson, Google and Amazon Web Services to C3.ai, to Dataiku, Tableau, R, NVIDIA, Accenture, and Heavy.AI). In the next 5-10 years society will experience the expansion from narrow AI to Artificial General Intelligence answering design and applications questions in a much broader sense (at least involving human and mobility user behavior, materials, manufacturing, mobility, and advanced computing sciences).
The next 5-to-10 years will see AI becoming a tool, a partner, and a method to learn, make decisions, and operate in our daily lives.
How can we advance AI?
Probably, the three areas with the most potential for advancement include industry-wide open-source data exchange formats, multi-company data integration, and partnering with academia to advance the topic of learning statistics, and image- and text-based AI technologies.
Currently what is the biggest challenge in AI?
Based on what I know today, the lack of examples of successful application of AI to braking system development, the proprietary nature of multiple sources of data, and how to merge the current way of developing systems with the new technologies which AI enables.
What are the most critical changes that we must make to face the future effectively?
Probably, accept that AI will get the engineering knowledge to a point where the answer is not fully understandable by humans, and embed tools and methods which pursue developments without the scientific and engineering knowledge of all the mechanisms behind the cause-and-effect relationships, as we perceive them. Corrosion, fuzzy tribology and wear, and environmental interactions are just three examples. Another aspect is to connect AI with the lifecycle of the system and its components. From well to wheel, and from concept to usage by the consumer.
When discussing this topic with others in the industry, what is the question about it that you are most frequently asked? How do you answer it?
There are two at least: how to select the proper suite of platforms, tools, and partners, and how to define a project which can benefit from AI and demonstrate its value to the company in the near term. This early success would secure funding and technical resources for expanding (AI) to multiple developments, and make it ubiquitous in the engineering processes for future vehicles. Automated driving systems have been at the forefront of AI developments. They are our ‘cousins and neighbors’ for us to reach out for expertise and methods.
Does one existential question remain as to what to do when the AI outcome contradicts the current knowledge and the experts in the field?
Comments