transferlearning.dev
#Transferlearning Development
#Solinftec | Agricultural Platform powered by AI | Integrated from pre-planting to post-harvesting | Algorithms reducing time spent driving and idling | Route optimization | Machine learning to automate operations
#LangChain | Generative AI ecosystem | Comparison View | Hotkeys in Annotation Queue | Tools to evaluate response quality: correctness | Pinecone
#Pulumi | generative AI ecosystem | Pinecone
#Vercel | generative AI ecosystem | Pinecone
#Pinecone | Serverless vector database architecture
#Anthropic | generative AI ecosystem | Pinecone
#Field AI | Autonomous systems for machines deployed to fields | Off-road autonomy | Field Foundation Model (FFM) | Enabling on-edge robotic autonomy | Enabling robots to conquer unstructured settings
#AI Port | AI Research | Port of Rotterdam
#Airwayz | AI based unmanned aircraft traffic management (UTM) system
#Awake.Al | Digital twin for Port of Rotterdam
#picknik.ai | Remote Robot Control | Model based design approach of Behavior Trees
#artificialanalysis.ai | Independent analysis of AI models and hosting providers
#Croq | LPU Inference Engine | Fast inference for computationally intensive applications with a sequential component | Better performance on LLMs compared to GPUs | Support PyTorch, TensorFlow, and ONNX for inference | GroqCloud is powered by scaled network of Language Processing Units | HW: crack, node, card
#Anyscale | generative AI ecosystem | Pinecone
#Cohere | generative AI ecosystem | Pinecone
#Confluent | generative AI ecosystem | Pinecone
#SEA.AI | Detecting floating objects early | Using thermal and optical cameras to catch also objects escaping conventional systems such as Radar or AIS: Unsignalled crafts or other floating obstacles, e.g., containers, tree trunks, buoys, inflatables, kayaks, persons over board | System computes input from lowlight and thermal cameras, using Machine Vision technology, deep learning capabilities and proprietary database of millions of annotated marine objects | High-resolution lowlight and thermal cameras | Real-time learning of water surface patterns | Searching for anomalies | Distinguishing water from non-water | Comparing anomalies with neural network | Recognize objects by matching combination of filters | Augmented reality video stream combined with map view | Intelligent alarming based on threat level | Detecting persons in water | On-board cameras with integrated image processing | Providing digital understanding of vessel surroundings on water | SEA.AI App on smartphone or tablet
#Catepillar | Cat Command for Excavating | Remote control system | Cat Command station | Non line of sight remote control | Improving machine productivity | Fast restart of operations | Virtual cab with familiar controls and machine displays | Up to five different machines controlled by one user | Access to machine setting just like in the machine | Solution to workforce shortages | Training advantages for newer operators | Opportunities to those with physical limitations | Attracts a new generation of workers | Increases the longevity of experienced operators
#Robotion | Industrial Robots
#RNA Automation | Vision systems | Automated feeding; handling, robotic systems
#Kinova Robotics | Ultra lightweight professional robots
#ABB Robotics | Robotics | Machine automation | Robotics campus | Collaborative and industrial robot arms | Autonomous mobile robots (AMRs) | Innovation hub for robotics startups | Offering startups support to run proof of concepts, lend products and technology expertise and venture capital
#Boston Dynamics | Spot robot | Dynamic sensing | Dispatching agile mobile robots equipped with modern sensors to collect data on site | Radiation mapping autonomously | Gamma ray sensors | Personal dosimeters | Neutron detection | Alpha inspection with pancake sensors| Ability to withstand large amounts of radiation | Cabability to carry a variety of radiation detection payloads | Ongoing safe operation and eventual decommissioning of nuclear power plants | Machine learning algorithms | Ultrasonic inspections of rotating equipment with Fluke SV600 | Early detection of bearing failures before they lead to costly breakdowns | Operators can capture critical data without the need for additional inline sensors or manual inspections | Leica BLK ARC payload | Autonomous routines | Repeatable laser scans | Precise 3D data for tasks like factory design, equipment installation, and change management | Spot can identify and avoid obstacles requiring additional caution in dynamic environments, including carts, wires, and ladders | Spot Hardware Updates | Tablet Controller Pro | Ergonomic joysticks, longer battery life, and HDMI pass-through | Spot robit operates in temperatures up to 55°C | Core I/O antenna upgrade enhances range and reliability | AI Institute | Artificial intelligence (AI) | Robotics | Intelligent machines | Cognitive AI | Athletic AI | Organic hardware design
#Infonera | Open optical networking | Disaggregation of transponders and line systems | Coherent optical transceivers | Simplifying new technology integration with open APIs, standardized YANG data models, and compact modular platforms | Allowing to select ideal products and technologies for each network layer and domain
#Fetch Robotics | Just in time material delivery in manufacturing
#RGoRobotics | Perception Engine | AGV | AMR | Environments where machines and people operate in close proximity
#GAM | Robotic and servo gear | GPL Robotic Planetary | GCL Cycloidal | GSL Strain Wave | Precision Rack & Pinion | Linear Mount Products for Cartesian Robots | Manufacturing | IL, U.S.
#Cobot | Collaborative Robots | Universal Robotics (UR) | Techman | AUBO
#HELUKABEL | Cables, Wires and Accessories for Robotics | Sensor technology and data | Network and bus technology | Control and motor | Customer specific solutions
#Agility Robotics | Humanoid Robot Factory in Salem,Oregon | Digit | Bipedal robot | Moving totes and packages | Unloading trailers at warehouses
#Harmonic Drive | Precision servo actuators | Gearheads | Gear component sets
#Enabled Robotics | Intralogistics automation for the manufacturing industry | Picking of objects as well as their transportation | Mobile cobot | Collaborative robot
#Built Robotics | Autonomous solar piling robot | Robotic pile driver | RPD 35 surveys, distributes, drives, and records every pile for an all-in-one machine | Hammer working at 500 blows per minute | Sled carrying system | Pile plot plan is uploaded to the robot computer | Pile plan is automatically divided into sequences of piles | Each sequence is manually loaded into pile baskets | Pile baskets are forked onto sleds with telehandler | Autonomous piling task is initiated | Robot installs all the piles in its payload, then returns to its starting position to be reloaded
#Apptronik | Human centered robotic systems | Robotics systems operating in close proximity to humans | Astra upper body humanoid robot | Dreamer upper body humanoid developed by MEKA for UT Austin HCRL
#ANYbotics | ANYmal | Legged inspector robot | Industrial inspection robot | Read off value of analog gauges | Read out analog counters | Check state of lever-operated valves | Capture photo for manual analysis | Continuous bookkeeping of structural changes | Scanning and documenting 3d environment data as a digital twin | Workforce browser-based ANYbotics Field Operator application | Field Operator mode: enables to steer, set points of interest, visualize recorded facility maps, and create inspection missions | Mission Controller mode: enables to safely control single ANYmal, monitor fleet, and execute validated missions from thousands of kilometers away | Robot Administrator: unlocks protected functions for fleet-wide software upgrades
#Figure.ai | Speech-to-speech conversation | Onboard vision language model | AI-driven vision system powered by 6 onboard RGB cameras | Hands with 16 degrees of freedom and human-equivalent strength | General purpose humanoid | Robots to handle general tasks | Enabling robots to learn and interact with environment | Humanoids designed to caring for elderly or even cooking meals | Focusing primarily on industrial warehouse applications to start
#ForwardX Robotics | Increasing picking productivity by reducing wasted movement for workflow efficiency
#UiPath | Robotic Process Automation (RPA)
#Liebherr | Mining machine automation | Open and interoperable mine automation platforms | Autonomous haulage technology | Machine automation for excavators and dozers | Onboard Analytics
#Bobcat | Compact equipment | Loaders | Excavators | Compact tractors | Utility products | Telehandlers | Mowers
#SparkAI | Combining people and technology to resolve AI edge cases, false positives, exceptions encountered live in production | Launch & scale automation products faster | Inject augmented human cognition anywhere in production | Launch new products with imperfect AI | Bootstrap new models for evaluation | Stop training in the lab | Deploy product to customers sooner | Let SparkAI protect against edge cases | Derive deeper real world insights on where your model is struggling | All while product is live, generating revenue | Autonomous tractor | John Deere | REST API | Python SDK
#Krone | Autonomous farming systems | Manufacturer of agricultural machinery | Autonomous drive unit acting as smart system that can plow, cultivate, sow, mow, turn and swath
#Levatas | Builds end to end AI solutions, machine learning models, and human in the loop systems automating visual inspection | Teamed with Boston Dynamics
#run.ai (RUN AI) | Accelerating AI development with comprehensive life cycle support from concept to deployment | Software Toold for AI System Development and Deployment
#Forterra | AutoDrive | Driverless system | Autonomous vehicle system.| Vehicle-platform and payload agnostic | In-Vehicle, Remote or Garrison Oversight | Off-Road/On-Road | GPS-Denied Operation | Single-Vehicle Waypoint Route Navigation | Multi-Vehicle Convoys and Platooning | Static and Dynamic Obstacle Avoidance | Mission Re-Pathing | Retrotraverse and Reverse Platooning, with Trailers | Ouster lidar
#OndoSense | Radar distance sensor | Sensor software: integrated into control system or used for independent quality monitoring | Object detection | Distance measurement | Position control | Agriculture: reliable height control of the field sprayer | Mining industry | Transport & Logistics | Shipping & Offshore | Mechanical and plant engineering | Metal and steel industry | Energy sector | Harsh industrial environments | Dust & smoke: no influence | Rain & snow: no influence | Radar frequency: 122GHz | Opening angle: ±3° | Measuring range: 0.3 – 40 m | Measuring rate: up to 100Hz | Output rate: up to 10 ms / 100 Hz | Measurement accuracy; up to ±1mm | Measurement precision: ±1mm | Communication protocol: RS485; Profinet, other interfaces via gateway | Switching output: 3x push-pull (PNP/NPN) | Analogue output: Current interface (4 – 20 mA) | Protection class: IP67
#Heliogen | AI-controlled concentrating solar thermal technology | AI, cameras, advanced computer vision software precisely aligni array of small mirrors reflecting and concentrating sunlight on receiver tower | Receiver generates heat which is transferred to thermal energy storage | Providing steam heat up to 300 °C | Cameras installed at top of tower measure color intensity of sky as reflected in mirrors | By comparing intensities as seen from multiple cameras, system calculates mirror orientation and direction of beam, for real-time hyper-accurate tracking | AI technology for continuous micro-adjustments | System automatically adapts to atmospheric conditions | WiFi connects heliostats | Direct Steam Generating Receivers (DSGR) absorb concentrated sunlight and transmit energy to pressurized water within metal tubes | Manufacturing facility in Long Beach, California
#BlueRobotics | Natural feature navigation | SLAM navigation | Driving AGV manually around the site | Recording data from vehicle safety laser scanners | Cleaning map by removing any dynamic objects | Programming required routes for AGV | Autonomous Navigation Technology (ANT) | Automated Guided Vehicles (AGV)
#Ecorobotix | Plant-by-Plant AI detection and treatment softwareAlgorithms to detect and treat an extensive array of crops and weeds | High-resolution camera system for scanning | Recognition and differentiation of useful and harmful plants by computer | Control over precision nozzels | Real-time imagery capturing amd object ifrntification in less than 250 milöiseconds | Spray footprint of 6x6 cm (2.4x2.4 in). | Precision agriculture by accurately detecting and treating individual plants
#Nordcloud | Aviation AI services
#Xwatch Safety Solutions | Safety solutions to machinery
#NVIDIA | AI Platform | Omniverse development platform | Universal Scene Description (OpenUSD) | Virtual Factory Reference Architecture | Factory digital twins | Robotics simulations for industrial manipulators, autonomous mobile robots, inspection cameras | Synthetic Data Generation
#BlinkLab (ASX: BB1) | Early diagnosis of autism spectrum disorder (ASD), attention deficit hyperactivity disorder and schizophrenia | Smartphone-neurobehavioral | High-speed video tracking of face | Mobile app as medical device to perform neurometric tests | Tests as biomarkers for neurological and psychiatric disorders | Smartphone camera and microphone as sensors | Results uploaded to cloud-based portal for storage and analysis using machine learning | Venture in digital healthcare
#Captain AI | Fully autonomous shipping solutions | AI models for object detection, image classification and instance segmentation | Creating comprehensive situational awareness | Route planner | Sensor fusion algorithm | Using radar, cameras, GPS, Octance, sonar and AIS installed on vessels | Three test vessels in the port of Rotterdam: tugboat from Kotug International, Watertaxi 14 from Watertaxi Rotterdam and patrol vessel from Port of Rotterdam Authority
#Stability AI | generative AI | Chatbot with text and image generation | Text-to-image technology | Multi-subject prompts | Model available via API | Generative AI video model | Music and sound effects generation with audio diffusion technology | 3D object generation from single images | Lenovo Creator Zone | Amazon Bedrock service from Amazon Web Services (AWS) | Stable Diffusion 3.5 | Multiple customizable variants | Run on consumer hardware
#Titomic | Titomic Kinetic Fusion Additive Manufacturing (TKF AM) systems | Integrating cold spray technologies, software and robotics into existing systems | Proprietary manufacturing cell | Manufacturing prototypes on a small scale | Cold spray powders | Rapid repairs | Metal coatings without heat | Build and incorporate electronic sensors directly into parts | Industry 4.0 platform | Linux NAS server
#Optiscan Imaging (ASX:OIL) | Digital confocal laser endomicroscopic imaging system tailored for application in robotic surgery | Digital endomicroscopic hardware and software | Robot-compatible endomicroscopic imaging system | Robotic-assisted breast cancer surgery | Embedding its technology as a key component of oncological surgery workflows | Collabirates with Mayo Clinuc to create a robot-compatible endomicroscopic imaging system
#Plus | Autonomous driving software solutions | High performance automatic collision mitigation | Event data recorder | Full situational awareness where vehicle is located | AI-based autonomous driving software | Both left- and right-hand driving | Long range cameras continuously auto-calibrate
#IDS | Industrial image processing | 3D cameras | Digital twins can distinguish color | Higher reproducible Z-accuracy | Stereo cameras: 240 mm, 455 mm | RGB sensor | Distinguishing colored objects | Improved pattern contrast on objects at long distances | Z accuracy: 0.1 mm at 1 m object distance | SDK | AI based image processing web service | AI based image analysis | Time Of Flght (TOF) camera
#HyTerra | White hydrogen exploration | Natural hydrogen and helium.| Nemaha project in Kansas City and Wichita | White hydrogen: lower costs and a cleaner alternative than hydrogen created from gas or water | Amazon | Bill Gates | Airborne geophysical survey by HyTerra
#The United States Geological Service (USGS) | White hydrogen searchers have been looking in wrong places | Right geologies are now known, race is on | USGS to publish map | A mix of at least 15% H2 could be blended into natural gas pipelines | Dedicated hydrogen pipelines
#Artificial Analysis | Independent analysis of AI language models and API providers
#MIT | Autonomous Systems Development Facility (ASDF) | Enabling the development and testing of autonomy algorithms and capabilities | Prototyping and testing of ground-based, aerial, and undersea autonomous systems | Infrared sensors | Integrated motion capture system | Reflective tags on vehicles | Indoor positioning system generating GPS signals | Dry running systems indoors before major outdoor field tests | UAV copter with propeller with low acoustic signature | System for countering unmanned aircraft in urban environment
#Navy Research Laboratory (NRL) | Laboratory for Autonomous Systems Research (LASR) | Prototyping High Bay | Littoral High Bay | Desert High Bay | Tropical High Bay
#perplexity.ai | Perplexity | AI search tool crafting answers| Citing sources | Real-time access to the web | Generating follow up questions after every answer | Multimodal search: text, images, videos, graphs, charts, and visuals cues | Pro Search: asks for details, considers preferences, dives deeper, delivers pinpoint results
#Bota | Force-sensitive robots
#RGo | Robotics platforms
#Synapticon | Motion control
#Force | Force-torque sensing systems
#Intelligent Actuator Incorporated | Actuators
#Netzer Precision | Rotary encoders
#Protech | Carbon fiber parts
#NavVis | Digital Factory | Mobile mapping | Laser scanning | Reality capture technology | Comprehensive spatial data with 360 degree photos | Info to architecture, engineering, and construction professionals (AEC) | Simultaneous Localization and Mapping (SLAM) | Verification | Overcoming production challenges | Improving collaboration in construction projects | EU Taxonomy compliance | Dynamic handheld laser scanner, NavVis MLX for confined, smaller spaces or shorter, more frequent scanning on site
#Cohere | Retrieval Augmented Generation (RAG) at production scale | Command: write product descriptions, help draft emails
#Supermicro | Liquid cooling | Tesla Gigafactory
#xAI | Elon Musk | Creating AI capable of advanced mathematical reasoning | Grok: AI chatbot integrated with Twitter (X)
#Neura Robotics | Multi-purpose humanoid robot | Cognitive robotics | Cobot arm | Mobile robot | Mobile manipulator
#Omron Robotics | Cognitive robotics
#Safety Technologies Inc | Cognitive robotics
#Swislog | AutoStore | Robotized storage and order processing solution | Uses robots and bins to quickly process small parts orders | Enables direct stacking of bins on top of each other | System learns which products have a higher rotation, storing them on the top layer | If one robot breaks down, another takes over its tasks | Six robots use the same as a toaster | Robots deliver goods directly to pick stations | Clients: Sportamore (Sweden), Bio Partner (Switzerland), Linfox (Australia), RoboStores (UAE), Radwell (USA)
#Mitsubishi Electric | Manufacturing and selling electrical and electronic equipment for factory automation, information processing and communications, space development and satellite communications, energy, transportation, and building equipment | Incorporating Realtime technology into factory automation (FA) control system devices, programmable logic controllers (PLCs), servo motors, and computer numerical controllers (CNCs)
#Realtime Robotics | Robot workcell and runtime systems
#Bishop-Wisecarver | Signature Motion Linear Actuators (SMLA)
#Robot Operating System (ROS) | Set of software libraries and tools helping to build robot applications | Open source software development kit for robotics applications | Instals on Ubuntu and Windows 10 | Software ecosystem of software for robotics | Linux, Windows, macOS | Embedded platforms (via micro-ROS) | Open Source Robotics Alliance (OSRA)
#Catepillar | Backup and prime power for data centres globally | Demand in part by cloud computing and generative artificial intelligence | Industrial Generator Sets | Diesel Generator Sets | Gas Generator Sets | Mobile Generator Sets
#John Cockerill | Galvatek | Surface treatment solutions | Automated surface treatment facilities delivery Chemical cleaning equipment for Maintenance, Repair and Overhaul (MRO) of engine parts in aviation industry | Effluent (wastewater) and water treatment plants | Automated solutions for aviation industry | MIL and ISO standards for defence | Automated gas turbine parts stripping lines for energy industry | Anodizing Plants | Chemical Cleaning Lines (CCL) | Chemical Surface Treatment Plants | Electroplating Plants | Chemical Stripping Lines
#Movu Robotics | Robotics | Automated pallet storage solution | Accessible cold storage automation | Temperature-controlled food and pharmaceutical industry | Space optimisation in storage | Streamlined materials handling process | Energy efficiency for cold storage | Robotics for low temperatures with icy and slippery surfaces | Automated Storage and Retrieval System (AS/RS) | Automating picking | Fleet management software
#GITAI | Autonomous dual robotic arm system (S2) | Un-space servicing, assembly, and manufacturing | On-orbit services: maintenance, inspection, and life-extension operations for satellites | Repair, inspection, and life extension services for satellites
#LookOut | AI vision system | Synthesized data from charts, AIS, computer vision, and cloud fusing it into one 3D augmented reality view | Connects to existing boat display | Mountable camera system to the top of any boat | Lookout App for laptop, phone or tablet | Infrared vision | Night vision sensor | Spotting small vessels, floating debris, buoys, people in water | Blind spot detection | Backup camera | Temperature breaks, bird cluster locations, underwater structures for anglers | Camera streaming over WiFi to phones and tablets on the boat | Over-the-air (OTA) updates | Marine-grade water-proof enclosure | Integrated with satellite compass | National Marine Electronics Association (NMEA) communication standard interface | Multifunction Display (MFD) | Multi-core CPU driving augmented reality compute stack | ClearCloud service | NVIDIA RTX GPU for real-time computer vision | DockWa app
#GivernyAI | Generative AI technology seamlessly translating and adapting video content across languages
#SiLC | MR Eyeonic Vision System | Optimized for ranges up to 150 meters | Dual polarization
#Flyability | Drones for industrial inspection and analysis | Confined space inspection | Collision and crash resistant inspection drone | 3D mapping | Volumetric measurement | Inspections of cargo ships, bridges, ports, steel mills cement factories, liquid gas tanks, nuclear facilities, city wide underground sewage systems | Ouster lidar
#National Technical University of Athens | MariNeXt deep-learning framework detecting and identifying marine pollution | Sentinel-2 imagery | Detecting marine debris and oil spills on sea surface | Automated data collection and analysis across large spatial and temporal scales | Deep learning framework | Data augmentation techniques | Multi-scale convolutional attention network | Marine Debris and Oil Spill (MADOS) dataset | cuDNN-accelerated PyTorch framework | NVIDIA RTX A5000 GPUs | NVIDIA Academic Hardware Grant Program | AI framework produced promising predictive maps | Shortcomings: unbalanced dataset, marine water and oil spills are abundant, foam and natural organic material are less represented
#Google DeepMind Technologies Limited | Creating advanced AI models and applications | Artificial intelligence systems ALOHA Unleashed and DemoStart | Helping robots perform complex tasks that require dexterous movement | Two-armed manipulation tasks | Simulations to improve real-world performance on multi-fingered robotic hand | Helping robots learn from human demonstrations | Translating images to action | High level of dexterity in bi-arm manipulation | Robot has two hands that can be teleoperated for training and data-collection | Allowing robots to learn how to perform new tasks with fewer demonstrations | Collectung demonstration data by remotely operating robot behavior | Applying diffusion method | Predicting robot actions from random noise | Helpung robot learn from data | Collaborating with DemoStart | DemoStart is helping new robots acquire dexterous behaviors in simulation | Google collaborating with Shadow Robot
#Neptune Labs | neptune.ai | Tracking foundation model training | Model training | Reproducing experiments | Rolling back to the last working stage of model | Transferring models across domains and teams | Monitoring parallel training jobs | Tracking jobs operating on different compute clusters | Rapidly identifying and resolving model training issues | Workflow set up to handle the most common model training scenarios | Tool to organize deep learning experiments
#UCLA | AI model analyzing medical images of diseases | Deep-learning framework | SLice Integration by Vision Transformer (SLIViT) | Analyzing retinal scan, ultrasound video, CT, MRI | Identifying potential disease-risk biomarkers | Using novel pre-training and fine-tuning method | Relying on large, accessible public data sets | NVIDIA T4 GPUs, NVIDIA V100 Tensor Core GPUs, NVIDIA CUDA used to conduct research | SLIViT makes large-scale, accurate analysis realistic | Disease biomarkers help understand disease trajectory of patients | Tailoring treatment to patients based on biomarkers found through SLIVIT | Model largely pre-trained on datasets of 2D scans | Fine-tuning model on 3D scans | Transfer learned model to identify different disease biomarkers by fine-tuning on datasets consisting of imagery from very different modalities and organs | Trained on 2D retinal scans and then fine-tuned model on MRI of liver | Helping model with downstream learnings even though different imagery domains
#Linux Foundation | LF AI & Data | Fostering open source innovation in artificial intelligence and data | Open Platform for Enterprise AI (OPEA) | Creating flexible, scalable Generative AI systems | Promoting sustainable ecosystem for open source AI solutions | Simplifying the deployment of generative AI (GenAI) systems | Standardization of Retrieval-Augmented Generation (RAG) | Supporting Linux development and open-source software projects | Linux kernel | Linus Torvalds
#Tampere University | Pneumatic touchpad | Soft touchpad sensing force, area and location of contact without electricity | Device utilises pneumatic channels | Can be used in environments such as MRI machines | Soft robots | Rehabilitation aids | Touchpad does not need electricity | It uses pneumatic channels embedded in the device for detection | Made entirely of soft silicone | 32 channels that adapt to touch | Precise enough to recognise handwritten letters | Recognizes multiple simultaneous touches | Ideal for use in devices such as MRI machines | If cancer tumours are found during MRI scan, pneumatic robot can take biopsy while patient is being scanned | Pneumatic device can be used in strong radiation or conditions where even small spark of electricity would cause serious hazard
#ARM Institute | Manufacturing Innovation Institute (MII) | Funded by Office of Secretary of Defense | Part of Manufacturing USA Network
#Italian Trade Agency (ITA) | ARM (Advanced Robotics for Manufacturing) Institute member
#NVIDIA | GH200 Grace Hopper Superchip | Boosting TTFT in multiturn user interactions | Building contextual understanding of input sequence | Retrieval Augmented Generation (RAG) | Converting user prompt to tokens and then to highly dense vectors | Dot product operations | Building mathematical representation of the relationship between all tokens in prompt | Operations repeated across different layers of the model | Generating key-value cache (KV cache) | Reusing KV cache | KV offloaded from GPU memory to higher capacity and lower cost CPU memory | KV cache reloading back to GPU memory | KV cache offloading, eliminates need to recompute KV cache without holding up valuable GPU memory | Enabling multiple users to interact with the same content without recalculating KV cache for each new user | Speeding code generation in integrated development environments (IDEs) that have LLM capabilities, one or more developers can submit multiple prompts interacting with single code script over extended periods of time, offloading initial KV cache calculations onto CPU memory and then reloading it for subsequent interactions avoids repeated recalculations and saves valuable GPU and infrastructure resources | NVIDIA TensorRT-LLM | Llama 3 70B model running on server with NVIDIA H100 Tensor Core GPUs connected through PCIe to x86 host processor | Compelling deployment strategy for data centers
#Allen Institute for Artifical Intelligence | Robot planning precise action points to perform tasks accurately and reliably | Vision Language Model (VLM) controlling robot behavior | Introducing automatic synthetic data generation pipeline | Instruction-tuning VLM to robotic domains and needs | Predicting image keypoint affordances given language instructions | RGB image rendered from procedurally generated 3D scene | Computing spatial relations from camera perspective | Generating affordances by sampling points within object masks and object-surface intersections | Instruction-point pairs fine-tune language model | RoboPoint predicts 2D action points from image and instruction, which are projected into 3D using depth map | Robot navigates to these 3D targets with motion planner | Combining object and space reference data with VQA and object detection data | Leveraging spatial reasoning, object detection, and affordance prediction from diverse sources | Enabling to generalize combinatorially.| Synthetic dataset used to teach RoboPoint relational object reference and free space reference | Red and ground boxes as visual prompts to indicate reference objects | Cyan dots as visualized ground truth | NVIDIA | | Universidad Catolica San Pablo | University of Washington
#Astera Labs | Intelligent Connectivity Platform | High-speed connectivity integrated circuits (ICs), modules, boards | Connectivity System Management and Optimization Software | Providing hyperscaler customers ability to deploy and operate high-performance AI and cloud infrastructure | Software-defined IC architecture | Distributed microcontrollers and sensors | Real-time link and device monitoring | Smart Fabric Switches | PCIe 6 fabric switch | Delivering maximum predictable performance to increase GPU utilization | Protocol and performance optimization for increased data transfer efficiency | Real-time telemetry data | Built-in protocol analyzer with link state history and timestamps | Full non-destructive eye scan for RX Lane margining | Self-test features to minimize link downtime and accelerate fault isolation | AI scale-up (GPU-to-GPU) | AI clusters | Generative AI applications | Nvidia supplier
#naptha.ai | Platform designed for building decentralized multi-agent systems | Python SDK to integrate of decentralized capabilities into existing systems | Naptha Hub provides reusable components
#Neurala | Visual Inspection Automation (VIA) software | Vision AI technology | Lifelong-Deep Neural Network | Helping manufacturers improve their vision inspection and quality control processes to increase productivity | Manufacturers can build better performing AI models that continuously improve processes and production efficiencies
#Anthropic | AI safety and research | AI powered software engineering | AI to turn market trends into marketing campaign | AI powered connection to codebases for help on software troubleshooting errors and identifying code optimizations | AI powered computer use | AI prompt caching | Syncing GitHub repositories with AI model | Researching how large language models work internally | Searching relations among data, compute, parameters, and performance of large-scale networks | Studying and engaging with safety issues of large models | Amazon Bedrock | Google Cloud Vertex | GitLab
#VORAGO | Radiation hardened and radiation tolerant microcontrollers and microprocessors | High-volume manufacturing to harden commercially designed semiconductor component | Empowers mission success | Radiation-hardened ICs that excel in extreme environments | ARM | Texas Instruments | Brainchip | Custom hardware and firmware solutions tailored to withstand extreme environments | Ensuring that mission-critical components remain resilient | Custom solutions, armed with space-grade electronics and an ARM Cortex microcontrollers | International Space Station: dies measuring the effects of protons and cosmic rays | Satelites: devices utilized in Department of Defense Space Test Program (STP) missions | CubeSats: ARM microvontroller | Semiconductor components and solutions for extreme temperature environments up to +200°C
#NavVis | Point cloud processing workflow | Surveying | Laser scanning | Process laser scan data captured using NavVis technology from anywhere with internet connection | View and validate every control point | Ensure data privacy with fully automated and integrated blurring functionality | Point cloud presets | Environment-specific point cloud modes | Generate photorealistic point clouds automatically cleaned of dynamic objects while preserving original details and colors | Upload your geo-referenced control points file in global coordinate system | Automatically geo-register your point cloud | View control points exact location on quality map | Locate, verify, select/deselect control points to identify any potential errors before processing begins | Automated image anonymization process | Detecting and blurring individuals’ faces, bodies, and license plates in images and point clouds captured with NavVis devices | Confidently create and share projects that meet strict data confidentiality | Meet compliance standards, including GDPR requirements | Download comprehensive quality report | Detailed information on data accuracy
#Robotics & AI Institute | Collaborates with Boston Dynamics | Developed jointly Reinforcement Learning Researcher Kit for Spot quadruped robot | Developing sim-to-real for mobility | Transferring simulation results to real robotic hardware | Bridging sim-to-reality gap | Training policies generating a variety of agile behavior on physical hardware | Trying to achieve novel, robust, and practical locomotion behavior | Improving whole body loco-manipulation | Developing robot capability to manipulate objects and fixtures, such as doors and levers, in conjunction with locomotion significantly enhancing its utility | Exploring new policies to improve robustness in scenarios | Exploring full-body contact strategies | Exploring high-performance, whole-body locomotion and tasks that require full-body contact strategies, such as dynamic running and full-body manipulation of heavy objects, necessitating close coordination between arms and legs | Aiming to utilize reinforcement learning to generate behavior during complex contact events without imposing strict requirements | Develop technology that enables future generations of intelligent machines | Streamlining processes for robots to achieve new skills | Developing perception, situational understanding, reasoning, cognitive functions underpinning robot abilities and combining them with advances in their physical capabilities | Conducting research in four core areas: cognitive AI, athletic AI, organic hardware design, and ethics related to robotics
#UC Berkeley, CA, USA | Professor Trevor Darrell | Advancing machine intelligence | Methods for training vision models | Enabling robots to determine appropriate actions in novel situations | Approaches to make VLMs smaller and more efficient while retaining accuracy | How LLMs can be used as visual reasoning coordinators, overseeing the use of multiple task-specific models | Utilizing visual intelligence at home while preserving privacy | Focused on advancements in object detection, semantic segmentation and feature extraction techniques | Researched advanced unsupervised learning techniques and adaptive models | Researched cross-modal methods that integrate various data types | Advised SafelyYou, Nexar, SuperAnnotate. Pinterest, Tyzx, IQ Engines, Koozoo, BotSquare/Flutter, MetaMind, Trendage, Center Stage, KiwiBot, WaveOne, DeepScale, Grabango | Co-founder and President of Prompt AI
#Trossen Robotics | Pi Zero (π0) | Open-source vision-language-action model | Designed for general robotic control | Zero-shot learning | Dexterous manipulation | Aloha Kit | Single policy capable of controlling multiple types of robots without retraining | Generalist robotic learning | Pi Zero was trained on diverse robots | Pi Zero was transferred seamlessly to bimanual Aloha platform | Pi Zero executed actions in a zero-shot setting without additional fine-tuning | Pi Zero run on standard computational resources | Hardware: 12th Gen Intel(R) Core(TM) i9-12950HX | NVIDIA RTX A4500 16G | RAM 64G | OS: Ubuntu 22.04 | Dependencies: PyTorch, CUDA, Docker | PaliGemma | Pre-trained Vision-Language Model (VLM) | PaliGemma allows Pi Zero to understand scenes and follow natural language instructions | Image Encoding: Vision Transformer (ViT) to process robot camera feeds | Text Encoding: Converts natural language commands into numerical representation | Fusion: Aligns image features and text embeddings, helping model determine which objects are relevant to task | Pi Zero learns smooth motion trajectories using Flow Matching | Pi Zero learns a velocity field to model how actions should evolve over time | Pi Zero generates entire sequences of movement | Pi Zero predicts multiple future actions in one go | Pi Zero executes actions in chunks | ROS Robot Arms | Aloha Solo package | Intel RealSense cameras | Compact tripod mount | Tripod overhead camera | Ubuntu 22.04 LTS