Wildfire Detection
IoT Mesh Network
WildFire generates 20% of global C02 emission and led to 140 billion USD economic loss just in 2024. Dryad Networks is a Berlin-based startup building the most advanced and affordable ultra-early wildfire detection system to tackle this problem.
I was originally approached by Dryad’s CTO as a consultant to identify opportunities to increase the speed and accuracy of the sales to deployment process. This effort impacted their product line and processes, and as a result reduced the quoting and planning time by 90 percent (from months to hours) with an accuracy of close to 100 percent.
Following this preliminary work, I joined the company as Head of Design. I now lead the design of hardware, software, and integrated tools, while contributing to the company’s product vision and strategy. My focus is on increasing our market share by making our AI-enabled wildfire detection network faster to deploy, easier to operate, and trusted by those who depend on it.
Key Stakeholders
Benoit Vitoux (Head of Design: UX UI & Industrial Design), Arthur Annebicque (Lead PM), Hannes Breul (Lead Front End), AncutaMorarasu (Front end).
Silvanet Platfom
Product Design
The Silvanet Web Platform helps users (Reseller and end customers)to plan, deploy, monitor and manage wildfire detection emergencies.It’s designed to support resellers during pilots and give customers ongoing value from their deployments. It acts as a an operational tool as well as a demonstration of our product capabilities preparing for potential integration via our API for largescale customers.
Planning made simple
Guide the user to design a site protection in minutes and not weeks ensuring Lorawan notwork capabilities with automated placement.
Autonomous deployment
Resellers can run full pilots from planning to fire testing with minimal support, making the service scalable.
Smart Alerts
Adaptive protection modes adjust detection sensitivity based on the site environment stopping false alarm.
Site insights
Fire risk dashboards and environmental data help customers track conditions and plan ahead.
Silvanet Estimator
AI Built Internal tool
No matter how easy our plannign tool was, our sales team needed a tool to create rapid but realistic device estimations based ona terrain index calculated with a .kml file uploaded by the user.
To move fast, I built and launched an internal product 100% coded with AI. I used a mix of Claude, UX pilot and Lovable, using a custom version of Shoelace with our design system as reference and a developper MapTiler API.
The result is accurate at 95%, impacting CAC and time positively as an exact planning is not required to provide a good quote ball park and assess customer true interest.
Deployment App
iOS and Android app design
The deployment app guide our customers or installer teams to deploy their mesh. Directly linked to our planning tool, the challenge was to build a tool simple enough to be used by non-trained users to deploy and test the mesh network successfully.
Silvanet Sensor
Industrial Design and Production.
I designed the fourth generation of wildfire Sensors. This has been a production challenge to integrate new sensors, RF antenna while ensuring IP67 and maintaining the production cost goals. This redesign improved the ergonomy and measured installation speed by ~25% thanks to a revised fixture design. The overall product language was aligned with the Border and Mesh Gateway to finally create a unified product line.
I was directly working on optimizing the design for mass production (600k Units year) with the shell plastic producer in Austria and assembly supplier in Germany.
Silvaguard
AI Drone System Interface Design.
The X-prize is a grand, incentive-based contest designed by the XPRIZE Foundation to address humanity's most significant global challenges. We are participating on the Wild fire section creating the first system able to identify and suppress a fire at scale.
My challenge was to build a user interface leveraging our wildfire sensor network to control a set of AI enabled drones which are automatically identifying the source of an alert, categorizing it and suppressing it.
After several rounds, we are now at the semi-final.
All images & video presented on this page are the propriety of Dryad Networks GmBH.