1 / 12

Aleksey (Alex) Gurevich CTO DXC Analytics (Americas) Lee Kedrie Account Chief Technologist

In-Vehicle Edge Analytics. Aleksey (Alex) Gurevich CTO DXC Analytics (Americas) Lee Kedrie Account Chief Technologist Automotive Industry Hewlett Packard Enterprise. ADAS – Advanced Driver Assistance Systems. The Data Challenge

bcastillo
Download Presentation

Aleksey (Alex) Gurevich CTO DXC Analytics (Americas) Lee Kedrie Account Chief Technologist

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. In-Vehicle Edge Analytics Aleksey (Alex) Gurevich CTO DXC Analytics (Americas) Lee Kedrie Account Chief Technologist Automotive Industry Hewlett Packard Enterprise

  2. ADAS – Advanced Driver Assistance Systems • The Data Challenge • Numerous streaming parallel data feeds (Sensors, vehicle systems, interconnected sources, etc) • Large Data sets ( ex: multiple High Def Video can generate more than 2.5 GB per second) • Difficult to store collected data in-vehicle (upwards of 50-100TB per day) • Difficult to offload data (Upwards of 50 - 100GBps required) • Limited ability to pre-process data • Multiple Data Formats (ex: ROSbag, ADTF, MDF4, ascii, csv, etc) • For Operations - Most Data needs stream analytics processing and can be discarded , retaining anomalies only. • For R&D Most data is not relevant, only anomalies and specific condition data needed (especially after more iterations and maturity of the system.)

  3. ADAS – Advanced Driver Assistance Systems • The Hardware Challenge • Component size – space is at a premium • Rugged format – Temperature variations (high and low), Vibration, humidity, etc… • Power consumption – Power is limited and DC format • Interconnectivity with other hardware modules

  4. ADAS – Advanced Driver Assistance Systems • Current state for R&D • Most R&D efforts very dependent on central data consolidation for analytics model development. • Challenge to collect the needed high volumes of data on vehicle (upwards of 50-100TB per day) • Narrow scope Niche solution components ( ex: Purpose built Logger devices with very limited scope and scale) • Limited to No on-board data pre-processing and analytics

  5. ADAS – Advanced Driver Assistance Systems • Current State Commercial • Connected car limited to mostly infotainment • ADAS challenged with narrow scope niche solutions • Volumes and speeds are increasing • Limited interconnectivity between ADAS vehicles

  6. DXC Robotic Drive Accelerators Accelerated Time to Analyze Through Intelligent Tagging, Data Management & Access to Relevant Data Data Centers R&D Engineers DataAnalysts Remote Work Platform Data Scientists 6 Robotic Drive Trainer 3 Robotic Drive Trainer 4 5 Geographic Distributed Data Lake Robotic Drive Analyzer Annotation Teams 3 Robotic Drive Trainer 1 2 3 7 Instant Access Robotic Drive Ingestor Robotic DriveOptimizer Robotic Drive Actor Car OEMs Suppliers Consumers

  7. Next Generation Solutions NOW - Converged Edge Analytics the ability to ingest data, analyze the information and take action at the edge, directly at the data source. The HPE EdgeLine solution option leverages a modular in-vehicle compatible mini-rack configuration that can satisfy the data collection and extraction needs (including direct connect and data logistics options), The HPE EdgeLine also has the ability to scale the storage and vary the commute scaling to accommodate broader operation use cases and flexible data management options including prioritized data offload via LTE and wireless, in-vehicle advanced analytics such as machine learning and data analysis. The scalable CPU, memory and disk allows for this to act a converged in-vehicle portable analytics platform that would allow for future in-vehicle analytics and data operations to further accelerate time to value and reduce associated costs and data transfer volumes.

  8. High Performance: SFF Hot-Swappable SAS SSD disks High Performance Compute Blades (ex: Dakota 5218T 16core 2.2GHz SFF 2933 DDR4 768 GB RAM) Compact Format: 221mm (8.7”) wide 219.2mm (8.63”) tall 431mm (16.9”) deep Edgeline – HPE Modular Converged Edge System NEBS Compliant 105W High Speed LAN (100Gbps) 4G/5G LTE Wi-Fi Enabled Ethernet Lan connections Parallel Write: Multiple Sensor ingest (ex: Radar, Lidar, Video each upwards of 20Gbps) Full Security/Monitoring Compatible with core security and monitoring solutions and remote Management Flexible Integration: Compatible with Linux. Can be connected as an extended name space of global analytics platform

  9. In-Vehicle Equipment Converged Edge System IN-Vehicle component Systems (ex: brakes) Driver UI SDC

  10. Beyond ADAS– Interconnected Vehicle Connected Vehicle Smart Cities Robotic Drive + Analytics Services = AI Solution Platform Robotic Drive Solution components Edge Analytics Ingestor Access Actor Training Optimizer Analyzer Trainer Geo-Distributed Data lake

  11. Cross Industry Alignment– Bus / Shuttle Train Transport Smart City Commercial Truck Cargo Trains Heavy-machine vehicles - Farming, Mining, Seaport. Planes Vertical Take-Off/Landing (VTOL), Helicopter Commercial Delivery / Military Drone, Beyond Visual Line of Sight (BVLOS) Drone Automated guided vehicles (AGVs) , Unmanned ground vehicles (UGVs) Cargo Ship / Boat / water surface vehicle Autonomous Underwater vehicle (AUV) Remote Operations: Remote Oil Rigs, Sea Drilling Smart Manufacturing Facilities Robotic Drive + Analytics Services = AI Solution Platform Robotic Drive Solution components Edge Analytics Ingestor Access Actor Training Optimizer Analyzer Trainer Geo-Distributed Data lake

  12. Thank you

More Related