Last week, we were joined by the innovative team from Nantoo, a company developing sustainable solutions for green space maintenance. The webinar, “Optimizing Home Appliance Design with Cloud-Native Simulation and Physics AI,” offered a deep dive into how Nantoo leveraged cloud-native simulation to overcome significant product development hurdles and how Physics AI is set to revolutionize this process even further.
The session featured insights from our co-founder and Product Manager, Alex Fischer, alongside Nantoo’s CEO and Founder, Beatrice Sileno , and Chief R&D Engineer, Andrea Taurino. They walked us through Nantoo’s journey of developing a multi-action system for leaf collection and how simulation was the key to their success.
For those who missed it, here are our top five highlights from the session.
On-Demand Webinar
If the highlights caught your interest, there are many more to see. Watch the on-demand Simulation Expert Series webinar from SimScale on how real-time simulation with AI is driving faster design cycles and superior products by clicking the link below.
1. The Challenge: From a Vision to a Scalable Product
Nantoo set out to solve a common and frustrating problem: the inefficient and messy process of collecting autumn leaves. Their solution is an ambitious integrated system that not only vacuums and shreds leaves into a compostable bag but also functions as a blower and supports various accessories for total outdoor cleaning.
However, turning this brilliant idea into a scalable product proved to be a massive challenge. Beatrice Sileno, Nantoo’s CEO, shared their initial struggles, stating, “getting from vision to reality has been far tougher and more frustrating than I ever imagined and every prototype came with a high price tag, long delays and a constant echo: this is impossible”. This frustration with physical prototyping led them to reimagine their design process, and they discovered a paradigm shift with SimScale.
Key Takeaway:
Evaluate your physical prototyping process for bottlenecks; if costs and delays are high, adopting a digital twin approach early can de-risk development and prevent frustrating setbacks.
2. The Solution: A 3-Phase Digital Twin Strategy
Andrea Taurino, Nantoo’s Chief R&D Engineer, detailed the company’s shift to a “digital twin” methodology, breaking down the complex design process into a manageable three-phase strategy. Instead of relying on costly physical prototypes, Nantoo embraced simulation to systematically optimize their product. The first phase focused on optimizing the core of the system, the impeller, using a “virtual wind tunnel” approach within SimScale to meet performance and low-power consumption targets. Once the impeller was optimized, the second phase shifted to the airflow within the complete machine to achieve perfect suction, blowing, and the cyclonic effect needed to keep leaves in the bag. Finally, the third phase used SimScale to develop and analyze various accessories, such as an electric broom and a flexible pipe, ensuring they integrated perfectly with the main unit.
Key Takeaway:
For complex product designs, break the project into manageable phases by first simulating and optimizing critical components in isolation before analyzing the complete system’s performance.
3. The Method: Smart Optimization with Taguchi
To avoid endless trial-and-error simulations, Nantoo employed the Taguchi method, a powerful statistical approach for design optimization. Andrea explained how they defined key control factors for the impeller—such as inner/outer diameters, blade shape, and twist—and used SimScale to analyze the design cases. This systematic approach required a significant number of simulations. For the impeller alone, Nantoo ran 13 iterations of the Taguchi method, totaling 247 simulations.
It would have definitely been impossible to do so many simulations cost effectively and rapidly without SimScale.
Andrea Taurino
The results were astounding: impeller efficiency in their test setup skyrocketed from an initial 20% to a remarkable 90%. This entire data-driven strategy was completed in just six months.
Key Takeaway:
Instead of manual trial-and-error, use statistical methods like the Taguchi approach combined with cloud computing to efficiently explore a vast design space and achieve significant performance gains.
4. The Future is Now: An Introduction to Physics AI
Building on the theme of rapid iteration, our co-founder Alex Fischer introduced our platform’s strategy for Physics AI. He explained that while traditional simulation has revolutionized engineering, Physics AI takes it a step further by dramatically cutting down the time to get results.
It works by feeding simulation results to a graph neural network, which is then trained to provide physics predictions in seconds. Alex demonstrated how AI models can be trained using two primary methods: running a targeted set of simulations on new, synthetic data specifically to train a model, or leveraging valuable existing data from past simulation projects to accelerate future designs.
Key Takeaway:
Leverage Physics AI to get near-instant feedback on design changes, making it feasible to run extensive optimization studies and test more ideas in a fraction of the time.
5. The Demo: Predicting Performance in Seconds
The highlight for many was the live demo, where Alex showed Physics AI in action. Using an AI model trained on synthetic leaf blower data, he predicted the performance of a completely new design variation in a matter of seconds—a process that would take about an hour with a traditional CFD simulation. Even more impressively, he used Nantoo’s own past simulation data to train a custom AI model on the fly. This model then accurately predicted the pressure field on an unseen impeller design, demonstrating how companies can build valuable, proprietary AI models from their existing engineering work. This capability allows engineers to use AI for rapid iteration to find the best design quickly, and then use a small number of traditional simulations for final validation.
Key Takeaway:
Treat your historical simulation data as a valuable asset; it can be used to train custom, proprietary AI models that accelerate future product development and build upon your team’s past work. Because Nantoo’s data on the SimScale platform was already organized and in the cloud it was ready to use for AI training with no additional work needed.
Final Thoughts
This webinar provided a compelling look at how cloud-native simulation empowers innovative companies like Nantoo to build better products faster. The integration of Physics AI promises to further accelerate this process, turning extensive simulation data into an invaluable asset for instant design feedback.
To get all the details and see the live demonstrations for yourself, be sure to watch the full webinar on demand!
