Persival GmbH’s Post

It can't be emphasized enough: Standardization initiatives such as ASAM OSI and ASAM OpenMATERIAL are key to making perception sensor simulation the development and testing tool it needs to be in the near future and we currently see massive breakthrougs🚀. We already need automated vehicles for e.g. public transport, logistics and agriculture and as Egon Wiedekind always says: The Game Changer Simulation will get us there!

Exciting Announcement from ISPRS Geospatial Week #gsw2023 Conference! I'm thrilled to introduce Mono-Hydra, our innovative extension to the Hydra framework. Venturing into the relatively uncharted domain of monocular camera setups for real-time spatial perception systems, Mono-Hydra bridges the gap between compact agility and real-time 3D scene graph generation. 🔍 Highlights: 1️⃣ Sub-20 cm error in real-time at 15 fps on NVIDIA 3080 GPU with a monocular camera setup. 2️⃣ Incorporates deep learning algorithms for depth and semantics predictions and a classical VIO algorithm - RVIO2. 🔗 Dive into our paper: https://lnkd.in/dxGf8aXj A huge shoutout to Prof. Francesco Nex and Prof. George Vosselman for their invaluable guidance. Gratitude to our team and the #GSW2023 community – let's keep pushing boundaries! 🚀 https://lnkd.in/d3tgPwyd #MonoHydra #uavcentre #UT #ITC #Robotics #SpatialPerception #Hydra #RVIO2 #DeepLearning

🌟 Sneak Peek Friday: Development in SMDL and Simspector! 🚗🌲 In this week's episode of Sneak Peek Fridays, we're excited to highlight a significant advancement in our sensor model development library (SMDL) and our data analysis and visualization tool, Simspector. Both can now work with alpha textures and normal maps! This enhancement is a game-changer for our simulation capabilities, providing both better performance and higher fidelity. Three months ago, we shared a video analyzing the 3D environment and moving objects for geometry (meshes/primitives) and material assignments. Now, with the integration of alpha textures and normal maps, we've taken a huge step forward in simulation detail. This improvement allows us to generate the lidar point cloud more accurately in complex environments, such as vegetation alongside the road. Watch the video below to see the impressive effects of these enhancements in action. The video showcases the exact same exemplary scene we created in-house, depicting a German highway, but now with significantly more detail on the vegetation. You'll notice a marked difference in the realism and precision of our simulations, showcasing the superior performance and fidelity we have achieved. We are thrilled with this development and look forward to continuing to push the boundaries of sensor simulation technology. Stay tuned for more updates and innovations from Persival!   #SneakPeekFridays #SMDL #Simspector #AlphaTextures #NormalMaps #Perception #Sensor #Simulation and #Model #Validation

What do virtual reality and augmented reality smart glasses, automobile Lidar, cell phone camera lenses, night vision, terrain mapping, and facial recognition have in common? All are technologies dramatically improved by metalenses—extremely thin optical structures that combine multiple functions of traditional and bulky curved optics into an ultracompact package. Now, development and adoption of these technologies is poised to accelerate substantially in Massachusetts, thanks to a $5 million award from the Massachusetts Technology Collaborative (MassTech), announced earlier this month. The grant will enable the University of Massachusetts Amherst to establish an open-access Advanced Optics Fabrication and Characterization Facility on its campus, a unique resource available to industry partners and researchers across the state. The MassTech grant also funds a major expansion of industry partner Electro Magnetic Applications, Inc.'s test and characterization capabilities to evaluate these new designs in real-world and harsh environments. Learn more: https://bit.ly/47a9FtF

Another example of why we us NavVis as one of our fundamental foundation tools for Digital Twins. Reach out if you're interested to learn more about how Digital Twins and BIM can benefit your next project.

#Mevea’s Senior Technical Advisor, Dr. Asko Rouvinen presented at the 2nd Autonomous Off-Highway Machinery Technology Summit (Feb 21-22, 2024) how advanced simulation, i.e. physics based #DigitalTwin can be used in assistive and autonomous system development and what benefits could be achieved. 👏😊 Two concrete examples presented were SLAM (Simultaneous Localisation and Mapping) application development and real-time simulation of structural flexibility. In the video the wheel loader with two LiDARs and odometer sensor is running in Unity visualization. Sensor signals utilize standard ROS (Robot Operating System) messages. LiDAR point cloud visualization includes deformable soil on ROS 3D Rviz visualizer. 💪 The link to the full article is at comments. 👍 #automation #OEM #AOMT2024 #construction

https://lnkd.in/gpJYMGRq exploring Omniverse isaac Ros simulation using urdf robot model. In this tutorial you will learn. 1.) Import urdf model into Omniverse environment. 2.)Add lidar sensor and camera. 3.)ROS bridge and visualise data in ros. 4.)Interface isaac ros with opencv. #nvidiaomniverse #nvidia

We're honored to be mentioned in the recent NVIDIA blog post "Accelerating Data Center Design With Digital Twins", which showcases how #rapidinnovation is made possible using digital twin technologies. The #realitycapture process starts with mobile lidar scanning using the revolutionary NavVis VLX 3 system. From there, Prevu3D was utilized to create an accurate 3D model in NVIDIA Omniverse. This enabled the virtual removal of the existing hardware and rapid design, optimization and installation of the new compute and network infrastructure - finding and fixing potential issues before they even occur! Come see us at booth 531 in the Industrial Digitalization Pavilion at #GTC24 to meet our service delivery partners NavVis and Prevu3D, and see these exciting technologies in action! #nvidiagtc #digitaltwin #digitaltwins #ai #artficialintelligence  #syntheticdata #scanning #laserscanning #slam #lidar #lidartechnology #industrial #industrialengineering #industrialai #realitycapture #computervision #nvidiaomniverse