The videos of the presentations and workshops from AnyLogic Conference 2019 are online. A diverse cross-section of organizations presented insightful and inspiring examples of simulation at work, highlighting the broad adoption of simulation modeling for commercial and industrial interests.
If you know what you want to see, jump straight to the playlist. Otherwise, here is a brief overview of the themes and some content pointers.
Different aspects of AI in simulation
Several presentations were related by their AI focus but provided very different insights. AnyLogic’s Arash Mahdavi, and Ty Wang of Skymind, explored the practical applications of deep reinforcement learning and demonstrated technologies for implementation. It is a great introduction to using simulation with machine learning, giving details of the platforms and APIs they use. You can explore the topic further with a documented version of the example model from the presentation.
Luigi Manca, Fair Dynamics Consulting, showed how reinforcement learning was used for decision making in a digital twin of a ferromagnetic core manufacturer. The presentation details some of the development challenges, demonstrating how the problem was broken down and why a combination of three neural networks were used instead of trying to find a single-network solution — Moving toward AI for optimal sequential decision making.
Away from the presentations about AI platforms and industrial applications, PwC’s Lyle Wallis develops the AI theme and re-imagines strategy decision making through simulation, gaming, AI, and continuous execution. The presentation looks at the management of rapid real-world change and deep uncertainty, with Lyle exploring how to develop robust strategies that minimize regret. An interesting aspect of the presentation concerns how insights and ideas can be communicated more effectively through storytelling and gamification.
Robots, and better decisions with digital twins
Robot automation offers improved profitability for grocery stores and a better customer experience. From initial idea and spreadsheet estimations, through simulation modeling evaluation, to real-world implementation, Amy Greer, of MOSIMTEC, and John Lert and Christian Hammel, of Alert Innovation, detail how Walmart put the technology to work and how the evaluation system has continued to develop. Modeling & Analysis of Grocery Picking Bots demonstrates the value of simulation modeling for decision making and how it can be adapted throughout the project lifecycle.
The continued use, and re-use, of simulation models was a key part of Goldratt Research Labs presentation. Dr. Alan Barnard and Jaco-Ben Vosloo demonstrated how a digital twin, constructed using their reusable project management system, helped a global gas-turbine manufacturer prioritize project tasks and maximize profitability. The presentation – Critical chain project management and simulation modeling – clearly demonstrates the utility of digital twin technology for prediction, helping inform management and enabling better decision making.
The presentations above are just a cross section of those given in Austin at the AnyLogic Conference 2019. You can find the whole set of videos, and accompanying presentation files, on the conference page. Photos from the event are also available in the follow-up.
It was great conference, thank you again to all who presented, attended and took part. For those unable to make it this year, we hope the videos can provide you with insight an inspiration.
Let us know in the comments below what you found most interesting, insightful, or inspiring!
⭐ Continuing the AI focus, join our Machine Learning Using Simulation webinar, July 11th.