Introduction
As I write from Singapore during a visit from Australia, the Grand Prix is less than five weeks away from hitting the streets. I’m a huge fan of the sport and when I saw the topic of artificial intelligence (AI) being discussed across multiple social platforms, I thought it was the perfect time to share an outsider’s view of AI’s potential to drive innovation in Formula One.
The race for AI dominance extends far beyond the racetrack as industries with operational assets strive to harness the power of AI for unprecedented gains in efficiency and sustainability.
At SIG Machine Learning, where I work, we recently reached an important milestone: the successful implementation of an AI-based control system for automated decision-making and control of multivariable, dynamic systems. This ‘intelligent’ control software is now deployed in the energy sector, saving time, reducing cost, and enhancing performance.
The application of AI in the energy sector has revealed its potential to revolutionise decision-making and automation in complex, data-rich environments—just like in Formula One.
Formula One parallels the energy sector. In both, simulations are used to enhance asset design and performance, remote control centres are used for real-time monitoring, on-site control systems ensure machinery safety and automation, and front-line teams apply technology to optimise efficiency and sustainability under time and cost pressures.
Across industries, we’re now able to combine AI modelling with engineering knowledge and control theory to deliver more intelligent, data-driven software, automating previously manual, repetitive decisions based on predetermined strategies.
Key Insight: Integrating domain knowledge into AI systems can improve agility and accuracy in decision-making across control, monitoring, and management. AI shines when it turns data into insights and automates formerly manual, time-consuming tasks.
In the energy sector, intelligent control systems oversee and coordinate the control of hundreds of assets in real-time. Built as a series of integrated modules, the software translates millions of data points into actionable insights and control system setpoint updates, using domain-knowledge-embedded machine learning algorithms to manage these complex, constantly changing systems.
Machine learning (ML), a subset of AI, is what is used to translate data into decisions that matter. We are using data-driven modelling to predict and optimise specific objectives within constraints, without needing to explicitly program every rule or sequence of steps to follow. These adaptive systems can handle complex, non-linear systems in seemingly unpredictable environments, learning from experience just as we do.
Having witnessed firsthand the transformative power of intelligent control software, I can’t help but think:
Could Formula One teams be leveraging intelligent control technology to drive value, both on and off the track?
The Race (to AI-Driven Operations) Is On
Artificial intelligence (AI) is transforming industries with operational assets, including Formula One, where engineering, science and technology collide.
Huge amounts of data are generated both on and off the track. As the industry evolves its data infrastructure, there is a growing opportunity to leverage data more directly in decision-making. From driver decisions on the track, to pitwall strategies and factory design, AI technology can drive performance optimisation across a racing organisation.
It’s no longer just about having the data, or a data and analytics platform. The focus must shift to investing in the intelligence of the applications that run on these platforms and embracing data-driven systems that can continue to learn and improve in performance over time, as we continue to operate.
AI-Driven Operations: The Power of Intelligent Control
AI-driven operations are built by applying AI to solve operational challenges. When AI is used to deliver solutions to problems, the software application could be considered an “intelligent application”.
Intelligent applications refer to software that combines domain knowledge with machine learning to deliver decision support (enriched insight to decision-makers) or intelligent control (autonomy with AI-driven control methods).
Intelligent control is one of the most valuable opportunities. By successfully embedding domain knowledge and strategic objectives into an automated system, we can not only improve the quality of decisions across teams, but the time and cost required to make those decisions.
This technology encodes engineering expertise and experience, into software, offering systematic, scalable, and transparent decision-making. The merging of Information Technology (IT) and Operational Technology (OT), enhanced computing power, and advanced machine learning, has created a new frontier for intelligent control systems.
Intelligent control technologies have been decades in the making.
Technology Innovation Timeline
Intelligent technologies are key to driving autonomy. Systems that can guide and make decisions (especially lower-level, repetitive, and manual decisions) contribute to more sustainable and efficient operations.
For teams looking to implement AI-driven operations to trial intelligent applications, introducing an autonomous operations program can be a practical way to explore opportunities to simplify and streamline operations.
Building an Effective Autonomous Operations Program
In this section, we’ll explore a hands-on approach to implementing AI.
In the image below, we’ve captured an example of how an Autonomous Operations Program can fit seamlessly into a digital transformation strategy.
In practice, we don’t know if the data available is sufficient to build a model that can outperform existing systems. For this reason, it’s valuable to partner with technology companies that have existing applications, or experience developing intelligent applications to understand the highest-value problems to solve.
Creating an Autonomous Operations Program
These programs can be built to directly target organisational value drivers. Building data, algorithms and models that target improvements in performance, cost or sustainability.
These programs also introduce a systematic way to capture learnings from race weekends. Feedback from drivers, the pitwall or remote teams can contribute post-race, to where the data-driven recommendations could be better defined or utilised.
This also brings people together with a common purpose – to translate thoughts, ideas and intuition into programmatic progress.
This is the exciting next chapter for motorsport. In the section below, I’ll outline an example of an Autonomous Operations Program, for teams looking to get started or continue driving improvements from AI-driven technologies.
Example: Autonomous Operations Program
The purpose of the program is to deliver value using AI-driven modelling. What does it look like to provide instructions to AI-driven systems? Use this data to optimise these objectives, under these constraints. Go.
Here’s a table of example initiatives to link specific value drivers with intelligent application trials.
Example Initiatives for an Autonomous Operations Program
Wrapping Up
The Formula One industry stands at an exciting threshold, with AI-driven technology poised to revolutionise the sport. As teams compete on the grid, data-driven insights hold the potential to dramatically enhance performance, decision-making, and overall racing strategies.
Drawing from my experience in the energy sector, where AI has transformed complex, data-rich environments, I’d suggest that intelligent control systems can significantly impact efficiency and outcomes in Formula One.
As we witness the convergence of data and systems, advancements in data processing and a focus on building intelligence, the race for AI dominance in Formula One is on. The future of the sport is data-driven, and the potential for innovation is boundless.
Thanks for reading. Please don’t hesitate to reach out with any questions.
LinkedIn: https://www.linkedin.com/in/samuelbost/
Comments