The Redhat® for ROS
Total Page:16
File Type:pdf, Size:1020Kb
The RedHat® for ROS So how come we don’t have robot maids? We will, once AGVs and Co-Bot arms reach the € 5k figure Till recently they were too expensive and dangerous Actually, we do have robot maids.. Almost every product we use has a CPU and some degree of autonomy and sensor feedback 100B ARM CPUs in circulation CPU market $300B “AI has replaced people in the financial services and robots have replaced labor in factories” Actually there’s much more to it 1999 2017 Travel agent Online reservation Cashier RFID, paypal, bitcoin, mobile.. Shopping, Supermarket Amazon 2 hour delivery, ebay Insurance agent, legal services Online auction News, journalist, encyclopedia Crowdsourced, AI generated Car industry Uber, driverless, shared Record, Video store Spotify, Netflix Consumer Voice NLP AI • Amazon Echo • Harmon Kardon Invoke • Google Home • Lenovo Smart Assistant • Google Home Max • Sonos One • Ultimate Ears Megablast • Apple HomePod • Baidu Family Robot • Tmall Genie GPU based products Hardware ANN • Nvidia Tesla • Nvidia Volta - with additional 'tensor units' accelerating calculations for neural networks • Radeon Instinct is AMD's line of GPU derived products for AI acceleration. AI accelerating co-processors • Snapdragon 845 contains a Hexagon 685 DSP core for AI processing in camera, voice, XR and gaming applications • Imagination Technologies - PowerVR 2NX NNA (Neural Net Accelerator) is an IP core to be licensed licensed • Neural Engine is an AI accelerator core within the Apple A11 Bionic SoC. • Cadence Tensilica Vision C5 is a neural networks optimized DSP IP core • HiSilicon Kirin 970 - Includes The Neural Processing Unit is a neural network accelerator • CEVA, Inc. – launched 4 AI processors called NeuPro. Programmable vector DSP + 8-bit or 16-bit neural network layers supporting neural nets with performances ranging from 2 TOPS thru 12.5 TOPS Research and unreleased products • Tesla Motors confirms a rumour that it is developing an AI chip for autonomous driving. • Eyeriss is an accelerator design aimed explicitly at convolutional neural networks • Kalray is an accelerator for convolutional neural nets. • SpiNNaker is a many-core design specialized for simulating a large neural network. • Graphcore IPU is a graph-based AI accelerator. • DPU, by wave computing, a dataflow architecture • STMicroelectronics - SoC manufactured in a 28 nm process containing a deep CNN accelerator. • TrueNorth is a manycore design based on spiking neurons rather than traditional arithmetic. • Intel Loihi is an experimental neuromorphic chip. • BrainChip - commercial PCI Express card with a Xilinx Kintex Ultrascale FPGA running neuromorphic neural cores applying pattern recognition on 600 video images per second using 16 watts of power. • IIT Madras is designing a spiking neuron accelerator for big-data analytics. Disrupting Industrial Automation Traditional automation • conveyor belts, cameras, pumps, motors, PLC • Robotic arms have been around for 40 years • Robotic AGVs have been around for 30 years • The car industry is symbiotic to the robot-arm industry • Industrial automation = MTBF, SLA, 10 year ROI, Integration Disruption • Recently more sensors, more computing power, collaborative • New robotic arms are sensitive enough to work without gate • New AGVs are smaller, able to stop in time, surpass obstacles Accelerated growth of robotics market Germany’s Industry 4.0 If we automate manufacturing and deploy good data analytics China will not have the advantage of $4/hr factory workers (assuming availability of supply chain) Made in China 2025 If we give incentives to adopt robots instead of workers in factories We will keep our upper hand as the world’s biggest manufacturer Aiming to have more than SK’s 550 robots per 10000 workers (and we already have the supply chain) Robot Arms • Universal Robots made a “killing” 6 years ago - $25k robot arm – Kukka and ABB charge $80k for 6 DOF robot arm – 2000 deployed – Sold for US$285 million in 2015 • Over 50 collaborative robot arms and startups in the last 3 years • Focus on small payload, slower speed, sensitive joints • The secret sauce of robot arms is the Harmonic Drive – Handful of manufacturers. $600 from China, $1500 from Germany/Japan • At least 10 Chinese Cobot companies – Foxcon deploy 100k robots a year replacing their 1.2m workforce – Estimated 3000 new robot companies the past 3 years in China • Robot Arms on Exponential growth – Franka startup planning to sell 8000 robot arms in 2018 – (ABB installed total of 300k robot arms) • 50 new co-bot arms in 4 years • Working safely around people • Price down to €10k (Harmonic driver bottleneck) • 60 new AGV startups the past 4 years • Using ROS opensource software • LIDAR, 3D camera, CPU/GPU AGV / Material Handling robots 100+ companies Locus Clearpath 6river Adept Fetch .. – Velodyne halved the price of its popular VLP-16 lidar sensor, from $8,000 to $4,000 – Several solid-state LIDAR startups launching $200 products Market SizeMobile robot market Market Current Forecasted In 5 years Material Handling (AGV) 1.71 2.6 TAM - $24B Cleaning 1.7 3 Robotic Automation Products Healthcare 1.7 2.8 Photo/Video* 1.5 4.0 Surveillance & Security 1.34 2.36 Telepresence .101 .503 Delivery .620 4 Retail .684 1.5 Cobot 1 3.3 Total Robotic Automation 10.4 24.1 Market 18.5% av. CAGR Note: Values are USD and in Billions *New Market – above shows the trend of an adjacent market - photography drones Sources: Markets and Markets Report, Drone Guru.co, World Robotics, Business Insider, business wire 100,000 Robot Developers use ROS (Robot Operating System – Open Source) MOV.AI THE FRAMEWORK FOR ROBOT AUTONOMY Fleet Management MOV.AI makes it easy to deploy Visual State Drag & Drop Logic Nodes, and maintain Pub/Sub lines complex ROS- based systems Cyber AI+HW Node Saves 50% of Security App-Store development time and cost BackOffice & Database End-User UI Integration Builder By 2025 10s of millions of people will loose their job pre-retirement Government of the future ? Collect more taxes Reduce Taxes by automating all public institutions Commercial freedom prevents brain-drain Stronger leadership Direct Democracy Because Everyone will be “unemployed” Aspire to regular career change and life-long learning Universal basic income Universal Basic Life Quality Automation will provide food, health, education, services AI to optimize public spending for citizen happiness .