Robert Rossi Email & Phone Number

San Diego , Ca, Us

As member of Executive Leadership Team reporting to the CEO actively engaged with C-Suite & Board of Directors on strategic matters & crisis management in addition to evolving the company’s L4 technology for Class 8 Autonomous Trucking under development:- Led all aspects of US Technology R&D and Fleet Operations including financial planning and control, organizational development and two successful organizational restructuring initiatives preparing the US Operations for a divestiture/acquisition.- Led organizational, technology development and operational aspects of attempted US Divestiture process including M&A pitch, Deep Dive Due Diligence, Class 8 L4 technology demonstrations and negotiations with prospect buyers.- Leveraged deep technical background and experience to lead “deep dive” problem solving of system deficits across full L4 autonomy stack, embedded hardware, onboard/offboard software, V&V etc.- Promoted TuSimple in Marketing events such as panel speaker at trade shows and social media blogs.

San Diego , Ca, Us

Led Product Unit delivering highly automated, human in the loop High-Definition Maps for L4 Autonomous Driving:- Led overall business and technical strategy, hiring, organizational development, and engineering execution [ program management, development management, testing/V&V and operations].- Led overarching Innovation and development for AWS hosted Map Construction and dynamic updating pipeline [K8/Argo/Postgres/Java/C++] - Led Operations and vehicle technology for small map data sourcing fleet.- Oversaw Operations for Map Content development team.

Amsterdam, North Holland, Nl

Led 350-person Engineering organization delivering HD Maps for automotive: Development of AWS, Cloud based, highly automated “human in the loop” High-Definition Map processing system for L4 Autonomous Driving and L2/L2+ ADAS Maps: ADAS Map | TomTom, HD Map | TomTom

San Jose, California, Us

Incubated platform for Autonomous Robots/Vehicles combining safety critical software design and formal methods for software verification with AI/Deep Learning methods. Developed wavelet/frame based adaptive sensor data processing framework, attention centric executive controller and system for training and validating reasoning framework.

San Francisco, Ca, Us

Led early stage company through initial deployment of a LiDAR based map creation, representation and localization framework to Ford Motor Co. Investigated and helped the team migrate Python code to C++ and evolve existing LiDAR specific Point Cloud registration and processing implementation/infrastructure, grid based localization scheme as well as human annotated digital cartography tools for building semantic maps.

Santa Cruz, California, Us

Led Seebright Wave from ideation and concept development thru fully functional prototype. Wave is a see though Augmented Reality headset that uses a smartphone as a processor and display source. Managed engineering staff, in-sourced engineering and led design and fabrication of optics, mechanical components, software and embedded electronics:Industrial Design:• Wave Headset and Hand Held Controller concept design w/external ID firm Engineering Design (3 iterations):• Optical Design ( Zemax and OSD SYNOPSIS ) - Non-uniform reflective optical design for display w/Steve Eckhardt (patent pending) - Periscope (Zemax)• PCB Design (Altium) -controller and inertial tracking add-on• Firmware – FreeRTOS, BlueTooth Low Energy• Mechanical Design (Solid works) of plastics (headset, opto-mechanical and controller)• Unity Game Engine SDK • Android and iOS C++ mobile software for Unity pluginsManufacturing/Design for Manufacturing:• 3D Printing of headset and controller plastic components via 3D Systems• Plastic Optics prototype - 3D Systems• Plastic Injection Mold Design for Optics and Headset ( w/various partners )

San Jose, California, Us

Founded XOTAR to explore the marriage of compression/streaming technology with robot perception, cognition and motor control. Initially innovated technology for Cloud Robotics which entails equipping autonomous robots with minimal electronics with wireless access to edge and back end processing. The company was largely R&D in this phase. During this time, I strengthened my knowledge of numeric methods and parallel matrix computations and studied decomposition and coordination of massively parallel computations on heterogeneous multi-core architectures. Also did much deeper investigations into computer vision approaches that benefited from wavelet and compressed sensing methods with a view towards computational dynamics, computer vision, multi-view image registration and large scale machine learning. Also studied differential geometric and lie algebraic methods for sensory motor coordination of vision and advanced robotic hands. Several prototype concepts of autonomous robots powered by remote cognition and perception were developed targeting applications in defense and agriculture. Contracted with Boeing for the development of camera and machine vision system for use in carbon fiber tape aircraft component (wing) layup (787 Dreamliner and 777x).

Redmond, Washington, Us

Group Program Manager of the Core Media Processing Team, a core Windows OS team that delivered DSP based components numerous divisions. The Core Media Processing Team developed media compression architectures that became standards, performed optimization for embedded consumer electronics systems and also published code in porting kits used by Silicon Chip/SoC Manufacturers to develop custom implementations of Windows Media technologies [ Digital Cameras/Video Recorders, Voice recorders, Speech Interfaces (Ford Sync )].Led all programs for image, video and audio ( speech and entertainment ) for all Microsoft platforms ( Windows Vista, Windows Live, XBox, Zune, Windows Automotive ). 15 person program management team guiding a 125 person engineering/R&D team. Position combined software engineering, business development, partner development (semiconductor and consumer electronics industry ) and project management with personal technical contributions to Stereo/3D imaging, scale-able video, image compression (JPEG XR) and audio. Worked with Intel, AMD, nVidia, ARM, ( and its ecosystem) on ISA's for media and signal processing and ASM programming for Vector Processing, VLIW, GPU etc.Led highly successful campaign to deliver the successor to the JPEG image coding format; JPEG XR which became an ISO standard in 2009. JPEG XR delivers very efficient compression for high dynamic range imagery and enables a host of new products/features in displays, digital cameras, digital image editing software and printers. Also led the deployment of the SMPTE standardized VC-1 High Definition Video CODEC with Hollywood Studios and post production studios & led the standardization of VC-1 in Blu-Ray Optical Media Format ( and HD-DVD ).

Redmond, Washington, Us

Hired on as a Senior Consultant to architect and develop technology related to mobile CRM in a strategic Alliance with Siebel (Oracle). Led the design and development of mobile CRM/Enterprise Social Networking technology for Microsoft and its partners leveraging instant messaging, speech, and innovative diagrammatic graphical user interfaces for device agnostic fluid/ad-hoc business processes (Tablet, SmartPhone, Desktop, Audio only Cell Phone).

Net Personix developed platform silicon architectures for low power speech driven wireless personal area/Bluetooth® networks. The company developed supporting technology for a cloud based speech driven multimodal mobile terminal (ala Apple SIRI). Speech compression was performed using a proprietary speaker dependent recognition design leveraging background noise suppression and voice articulation enhancement delivering a language specific phoneme stream to a cloud based natural language interpretation and application environment.

Bozeman, Montana, Us

Neuron Data initially developed an Expert System knowledge representation and reasoning system written in C ( as opposed to LISP) for use on personal computers called Nexpert. The company went on to develop cross platform GUI system (Open Interface), acquired distributed processing tech similar to OpenMP and OpenMPI and data access technology. ND became Blaze Software and developed EJB/J2EE based Advisor business rules/expert system development and deployment environment in the latter 90's.Originally working as an application architect in Houston where my focus was oil & gas exploration and drilling etc., I moved to NYC to build the company’s backbone in financial systems which became the foundation of their eventual IPO. I became an advanced C++ developer during the mid 90’s working with other members of the Sun Users Group and C++ quant talent in NYC and worked for almost all of the big banks and insurance companies. C++ was used to extend Nexpert for domain specific application frameworks driven by knowledge based inference, leveraging distributed shared memory technology to create blackboard reasoning frameworks where several expert systems reason about different aspects of a problem and share certain memory objects. I typically integrated Nexpert with other products for hybrid sub-symbolic reasoning (neural network and other machine learning technology). In the latter part of my career I developed advanced server frameworks for embedding expert systems in large scale online transaction systems using Tuxedo, Oracle and CORBA. I was an early advocate of development using C++ templates for large scale frameworks and was an innovator in what is now called Template Metaprogramming. Some major applications were Norwest Financial Systems/Wells Fargo ( 1 Billion Dollar project), Nationwide Insurance, NATO (Combat theater on Convex Supercomputers) . Neuron Data’s name changed to Blaze Software and IPO’d in 2000 and was then acquired by Fair Isaac Corp.

Los Angeles, California, Us

Inference corporation was a pioneering company in first generation AI/Expert Systems. It originally developed an OPS5/Rete algorithm based Expert System tool called the Automated Reasoning Tool (ART) which was written in Lisp. I developed in Lisp and ART and also used C programming for serial interfacing to factory computers. After joining the company, I influenced Inference's founder (late Chuck Williams) to develop the same type of technology in C for Microsoft DOS and the emerging Microsoft Windows. This product was called ART-IM. I worked in Dearborn MI. at Ford Motor Company's Operations Research facility and developed applications for diagnostic problem solving for advanced manufacturing systems ( specifically electro-hydraulic test bench systems ). This system verified the correct operation of all manufactured automatic transmission valve bodies which Ford experienced quality issues with at the time. Inference began working on Case Based Reasoning technology near the end of my tenure which was subsequently marketed for email filtering and CRM. The company went public in on NASDAQ in 1996 and merged with eGain a few years later.

Founded a company that developed low cost laser based machine vision sensors using the z80 microprocessor coupled to Fairchild CCD image sensors. Developed an 8 bit precision image processing library in z80 ASM for basic image processing operators and sensor calibration. Developed statistical processing software and diagnostic expert systems using early Lisp Machine based Symbolic AI tech. Developed early applications of diagnostic reasoning systems based on emerging theories of qualitative reasoning about physical systems. ISO also did development on VAX 11/780's ( in FORTRAN) which were popular in automotive factories as well as Motorolla 68XXX based Microprocessors. The company additionally developed sensor data processing systems ( now called sensor fusion ) based on VME bus based components and the VRTX embedded OS. We also developed adaptive weld controllers and experimented with factory communication networks with a prototype implementation of the MAP/TOP protocol ( this was before Ethernet had established dominance ).

Detroit, Michigan, Us

I joined General Motors at the age of 16 under a high school co-op program. Initially worked in a team that was pioneering the use of laser based machine vision for inspection. The division I worked for was Fisher Body Division which produced the bodies of all GM vehicles. I learned to program in FORTRAN (on IBM 360) learning numerical analysis, splines which were used for surface modelling of the car body panels, and algorithms for kinematics/dynamics (complex mechanical linkages and motion systems ) and also worked on scheduling systems for die production and numerous special assignments. Also learned to program and worked heavily with Allen Bradley Programmable Controllers which controlled everything that moved in Fisher Body's massive factories. Learned most processes in sheet metal fabrication and assembly, finishing, glass and interiors. Also learned quite a bit about engineering management and factory management of vehicles. GM provided an amazing start to my career and I am forever grateful to them.