Applicant:

 

Robert B. Stout                 11910 Sixth St. Houston, TX 77072                rbs@snippets.org

voice: (281) 495‑5533                 mobile: (281) 630‑6318                  rbs521@earthlink.net

 

Skills:

 

Specialty:  Technology/software startups& “blank paper” designs, "skunk works" development, smart sensors, embedded systems design and architecture, software process audits & QA assessments.

 

Software:  System software & real time industrial applications of microcomputers, microcrontrollers, and DSP chips. Languages: Assembly, C, C++, Perl, Python, Forth, BASIC, LabVIEW, Modula‑2 and others. Environments: PC/MS‑DOS, Unix (Solaris, Linux, HP‑UX, AIX, BSD), Windows 32, various RTOSs, & others.

 

Hardware:  Conceptual & detail design. Digital design of systems and subsystems from CMOS to ECL. Analog design of servo systems, filters, data acquisition and geophysical systems. System and network design,

 

Science:  Physics (acoustics, nuclear, optics, photonics, rheology), chemistry, digital signal analysis.

 

Management & supervisory: Technical and non‑technical personnel, Budgeting, scheduling, salary reviews, R&D and market planning, system/ network administration, policies and procedures. Working knowledge of ISO 9001, ISO 9000‑3, and CMM.

 

 

Work Experience:

 

Metrix Instrument Co.               Position: Software Engineering Consultant (contract through MicroFirm)

2007                                         Synopsis: Principle architect of new multiprocessor (ARM7 and ARM9) vibration measuring platform. Also responsible for all DSP coding.

Details: Designed two IIR filters (one 18-pole BP, one single pole LP integrator), an FIR anti-aliasing filter using Quickfilter QF1D512 chips, and a 6x decimation filter.

 

Emerson Process Management  Title/Position: Software Engineering Consultant (contract through TPI Staffing)

2007                                         Synopsis: Analyzed code to identify root causes of mathematical non-repeatability in new gas chromatograph. (C++, UML, automated analysis tools)

Details: Used a combination of automated analysis tools, visual review, and debugging to repair algorithms, refactor and restructure code, and correct coding errors. This project used embedded Linux in a multi-processor system running on PowerPC and ARM7.

 

RFTrax/Fairfield                        Title/Position: Research Scientist

2004-2006                                 Synopsis: Working with emerging sensor technologies for new homeland and cargo security applications (MSP430 & AVR C & assy, DSP, electronic logic & analog active elliptical filters, physics, chemistry, optics, photonics, acoustics, nuclear).

                                                Details: Started out as a contractor to Fairfield Industries researching sensors for intermodal cargo container security. Became the “go to” guy for all sensor-related issues. After Fairfield acquired RFTrax as a subsidiary continued working on sensors until my position was eliminated do to a revised focus on the low-tech rail car tracking market. All designs shared a requirement for very low (≤ 3 mW) average power consumption.

Projects:

Intelligent triaxial accelerometer:

·         Did the complete high-level design, hardware and firmware.

·         Did the detailed hardware design of the accelerometer’s micro power 4^th‑order analog anti-aliasing elliptical active filters (3) using two Microchip 600 nA dual opamps per channel.

·         Participated in the detailed design of the (3) analog track and hold circuits.

·         Designed and implemented an 84-tap FIR bandpass filter, implemented on a TI MSP430F1611.

·         Directed and participated in firmware development.

·         Wrote specialized LabVIEW test instrumentation for engineering and production test.

Next-generation intelligent triaxial accelerometer:

·         Did preliminary redesign work to further reduce cost, including replacing the analog anti-aliasing filters with SPI-connected external dedicated FIR filters from Quickfilter Technologies.

 

Other sensors:

·         Did research and some development work on a common spectrometric platform to be used for both gamma-ray spectrometry and chemical sensors using IR/vis/UV spectra and/or gas chromatography. The challenge was to provide sufficient computing horsepower for advanced pattern matching algorithms while keeping the average power consumption in our target range.

·         Did research and preliminary design work on several promising low-cost chemical spectrometric sensor devices.

·         Designed a common platform for simple electrochemical cell (chemical) sensors. Once the design was complete, further work was shelved awaiting commercial demand.

·         Participated in the re-specification, redesign, and performance testing of our CZT-based gamma-ray sensor. Testing was done at Southwest Research, with whom I was the primary contact.

·         Did preliminary design work on neutron sensors.

·         Did preliminary research and design work on advanced photonic sensors for chemical and biological warfare agents, various explosives and toxic industrial chemicals.

 

Pathfinder Energy Services        Title/Position: Contractor (through Technipower)

2003-2004                                 Synopsis: Firmware re-design and lead programmer of the Pathmaker 3-D Rotational Steering tool (68332 C & assy, trigonometry of rotational device).

                                                Details: Hired to rewrite the fragile Pathmaker firmware for maintainability. I started with a complete redesign. PFES didn’t want to use an RTOS, so I designed one into the firmware. I did the top-level coding and another firmware engineer assigned to the project did the low-level coding. I was also urged to help teach him good coding practices. The whole effort was an apparent success. I recently talked to Mike Moody, the project engineer, and he said that the new firmware was proving to be quite robust and resilient.

 

Motorola/Metrowerks                Title/Position: Contractor (through Technisource)

2001-2001                                 Synopsis: QA tool development for StarCore DSP products (C, assy, & perl; Solaris, Linux, & Win32; DSP).

                                                Details: Originally hired to be an extra pair of hands, I quickly became the QA department’s resident C language lawyer. When I saw that it was taking all weekend (when it didn’t fail!) to run a full regression test using the Plauger and Plum Hall test suites, I wrote a suite of test automation tools (Perl scripts and dedicated test tools in compiled C) which would run a full regression test in about 2 hours.

                                                Postscript: Metrowerks offered me a full-time position, but not enough salary to justify my relocating to Austin. We were still negotiating when 9/11 happened and all deals were off the table.

 

Additech                                   Title/Position: Director of technology.

1997-2001                                 Synopsis: Conceptual design and implementation of networked multimedia and real time control system (x86 & 8051 C & assy; DOS, Linux, & NTe).

                                                Details: Originally hired as a contractor (through Technipower) to evaluate the work of another contracting firm, I was forced to report that their system would never work as designed. Since they already had two years and several million dollars invested in it, that wasn’t what they wanted to hear. The president then asked me if it was my money, what would I do? I told him that I would continue to let the other guys work on it for a while (who knows, I may have been wrong, or they might stumble into some way of making it work), but that I would immediately start a parallel effort to replace everything. Given that it had already taken two years to get to that point, he was very discouraged, but asked how long that might take. I told him that it really wasn‘t rocket science and I should be able to replace the firmware while retaining or salvaging most of the hardware in about 6 weeks. He and the board of directors were somewhat incredulous, but told me to go ahead. Six weeks later, I had the first prototype built on a shoestring using mostly my own software tools. The first production field unit was subsequently installed in Birmingham, AL, and operational about 2 months later. I continued to maintain and develop the design through several more generations. The system included a multimedia control computer at each fuel dispenser running MS/PC-DOS (it’s all we could afford and we had lots of licensed copies from the original design), an 8051-based real-time interface to the fuel metering and control hardware at each fuel dispenser, and a single back office system which acted as a file server and network interface running Debian Linux on standard PC hardware. Communications between the real-time interface and the multimedia computer was via a proprietary serial protocol. Communications between the multimedia computer and the file server used a proprietary UDP-based protocol

 

Finally, when sales didn’t meet the board’s expectations, the top management was replaced. I stayed around and fought an ultimately futile battle to upgrade the design in sensible ways, but one of the new president’s cronies from a previous job came in and redesigned everything at a significantly greater cost. Ten months later, after they had some confidence in the new design and I was no longer needed for sustaining, I was let go. Subsequently, the new president and everyone he’d brought in with him were fired.

 

Compaq Computer                    Title/Position: Contractor (through Technipower)

1996-1997                                 Synopsis: Developed V.80 protocol for dedicated modem (68332 C, ITU standards).

                                                Details: In the late 90’s, Compaq was in the modem business. Their modem was based on some COTS code for the base protocols, to which other protocols had been added by many different people at different times with different skill levels. Writing the V.80 protocol wasn’t too challenging. Making it play nice with the existing mass of spaghetti code was a nightmare.

 

Interface Design                       Title/Position: Contractor (through MicroFirm)

1995-2001                                 Synopsis: Embedded spectrograph sampling/control systems (8051, Z80 C & assy; motion control).

                                                Details: I worked for ID on several occasions and multiple projects. In each case, I wrote firmware to control automated sampling and data acquisition for spectrographs. None of these assignments lasted more than a few weeks, but I did wind up doing sustaining engineering on them for six years until ID was sold.

 

Meditrol (now Pyxis)                 Title/Position: Contractor (through MicroFirm)

1995-1996                                 Synopsis: Drug dispensing system firmware (x86, 6800 C & Dataflex, motion control).

                                                Details: I got this job based on a referral from someone I’d worked with at Baker CAC. Meditrol built automated networked drug dispensing cabinets for use in hospitals, controlled by a central prescription computer. The mechanism was essentially the same as a candy vending machine, but with an internal elevator and locking door mechanism. The central computer was programmed in the Dataflex database programming language with which I had experience. I therefore wrote 6802 firmware for the cabinets, Dataflex software for the control computer, and DOS/Windows applications for testing.

 

Baker CAC                              Title/Position: Manager, firmware R&E.

(now eProduction Solutions)       Synopsis: Staff = 9 + consultants, budget = $561K (1993 dollars). Managed development

1990-1994                                 of 8800 RPC, 6800 RTU, Trans-Java Gas Pipeline firmware. Directly developed algorithms and protocols (x86 & NSC32xxx C; QNX, SMX, HP-UX, & DOS; polynomial curve fitting analysis, SPE & IGU standards).

                                                Details: During my tenure at CAC, I oversaw shutting down the development center at the Santa Fe Springs (Los Angeles area) California office. Most work was brought back to Houston, but I did retain one critical contractor in Tarzana, which got me in trouble with the brass. Later, when CAC lost a patent infringement case to Delta-X, that contractor was able to help CAC rewrite the firmware for its cash cow product while losing only a single day’s sales while the inventory was flushed. Although I wasn’t a “hands on” manager, I did do most of the high-level design work and occasional necessary ancillary tasks. In one of those tasks, the sales people had sold an RTU with the promise that it could process incoming data in accordance with an existing API standard. Unfortunately, the tables in the API standard were huge, empirically derived, and, if digitized, would take up more memory than the RTU could support. I had my people go ahead with everything else while I began working on a computational model to duplicate the table data to within 0.1% accuracy using a curve fitting algorithm. It took two tries, but I wound up with the customer’s blessing to use an algorithm that used six lines of C code. However, my most significant achievement at CAC was successfully managing 24 separate projects at one time. It nearly killed me (I was hospitalized with congestive heart failure a short time after I left) though and I wouldn’t try it again. Following a reorganization in 1993, I spent my last nine months there as manager of software QA, where I wrote the ISO-9000-3 manual. Finally, CAC sold the company to Thermo Electron and my position was eliminated.

 

Cray Research                          Position/Title: Contractor (through MicroFirm)

Los Alamos National Lab          Synopsis: Consulted on high speed channel development with Los Alamos (ECL logic).

1988-1989                                 Details: Cray had an internal channel called the HSC (High Speed Channel) used to communicate between different heads in the computer. Los Alamos wanted to bring the HSC to the outside world in order to connect it to a frame buffer to view CFD data in real time. I already had contacts within Cray and had helped them once before in hooking up an Arriflex movie camera to a connected workstation, so they asked me to work on it. Shortly thereafter, LANL decided they wanted a non-proprietary channel, so the HPPI (High Performance Parallel Interface) was born. I continued to work on it for a while and sat on the ANSI HPPI committee. After enough changes to the spec, I bowed out after completing a first ECL prototype. HPPI has since been superseded by various high-speed serial standards.

 

Intermedics                               Position/Title: Contractor (through MicroFirm)

1986-1988                                 Synopsis: Pacemaker & insulin pump programmers (6809, 6805 assy).

                                                Details: I began the relationship with Intermedics when I worked for McClelland (see below) since McClelland did contract development work for Intermedics test department to help cover the operating expenses of the special projects group. After leaving McClelland, I continued doing work for Intermedics, developing telemetry equipment and firmware. Among the things I developed for them:

·         Firmware to turn their pacemaker programmer into a programmer for an experimental implantable insulin pump. This was all written in 6809 assembly.

·         They liked the telemetry firmware I wrote for the insulin pump so much that they asked me to retrofit it to their pacemaker programmers.

·         I helped with one more project for their test department, but it was a pretty trivial assignment and I can no longer recall what it was.

 

Symantec (nee Zortech)            Position/Title: Consultant (through MicroFirm)

1990                                         Synopsis: Wrote portion of C/C++ run‑time library (x86 C & assy; DOS & Win16/32).

                                                Details: In the late 80’s, I wrote the MicroFirm Function Library (MFL), a large C utility library, originally for my own use, but later released as shareware. During this same time period, I was the moderator of the FidoNet C_Echo, one of the more influential C language groups in the pre-Internet era. I had long been a proponent of Walter Bright’s compilers, he having written Datalight C and Zortech C++ (the first C++ compiler for PCs). Walter became a friend and he asked to use some of my MFL functions in the Zortech compiler standard library. A short while later, Symantec decided it wanted to compete with Microsoft in the development tools business and bought Zortech. So, Walter’s compiler became the Symantec C++ compiler - still with my library functions. Over time, as DOS and Windows evolved, I had to maintain the MFL code through several generations. Symantec later dropped out of the C++ compiler business. The compiler, still with my library functions, is now available from Walter’s new company, Digital Mars.

 

Texas Elevator Co.                    Position/Title: Contractor (through MicroFirm)

1986-1990                                 Synopsis: Elevator control systems (Z80 assy; adaptive algorithms),

                                                Details: Referred by a friend, I designed private label elevator control systems for Texas Elevator. The system was, of necessity, written in Z80 assembly. It was quite challenging since it had to control all sorts of hidden operational modes (e.g. fireman’s service, maintenance, etc.) as well as handling intelligent car scheduling and traffic management. We wound up installing two systems while I was still working with them. Both worked fine after a few initial issues which never affected rider safety. To the best of my knowledge, they installed a couple of more systems after I left. Setup was quite simple and didn’t require any specific programming skills. To the best of my knowledge, Texas Elevator has since gone out of business, probably following the death of its owner.

 

McClelland Engineers                Position/Title: Head of electronic special projects group

1985-1986                                 Synopsis: Development of custom Geotechnical instrumentation (Doppler radar, analog active elliptical filters, signal analysis).

                                                Details: Specific projects I worked on included:

·         The Ram Velocity Monitor, a Doppler radar device to measure the impact velocity of pile driving hammers. When a hammer hits a pile, it sets up ringing in the hammer. If the speed of the hammer is set improperly, the next blow comes just as the previous impact wave reflects back to the hammer end. If this happens, all of the hammer’s energy is dissipated in trying to destroy the hammer. If the speed is set correctly, the impact waves in the hammer will reinforce the hammer impact rather than cancelling it. The RVM allowed an operator to monitor the progress of a pile driving operation and adjust the speed as necessary. In practice, it was quite easy for even an inexperienced user to interpret the output of the device on a strip chart recorder.

·         The SIometer, a road roughness meter designed in cooperation with the Texas Highway department and the University of Texas at Arlington. The SIometer used a complex regression algorithm developed at UTA to remove the effects of a test vehicle’s suspension from accelerometer data. Used to inventory roads and schedule repaving crews, the complete SIometer system consisted of a large electronics package and a trunk-mounted accelerometer package. In operation, the driver would first drive at a constant 50 MPH on a calibrated road. This calibration road was used to model the vehicle dynamics so that their effects could be removed during an actual test run. The device worked reasonably well, but the highway department ultimately decided to stick with their existing 5^th wheel testers.

·         Dedicated test instruments produced for Intermedics: the Tens unit tester. "TENS" is the acronym for Transcutaneous Electrical Nerve Stimulation. A "TENS unit" is a pocket size, portable, battery-operated device that sends electrical impulses to certain parts of the body to block pain signals. The Intermedics unit was capable of a complex assortment of amplitude and frequency modulations. A production test system had to test a rack full of these devices in an environmental chamber. It would have been an easy task for an FFT analyzer, but my predecessor had bid the job at $20k each, complete with cabling and test racks. I wound up using an Apple II mother board along with an oscilloscope plug-in and digital I/O boards to control power and wrote some 40k of assembly code to run the specific tests. To achieve the performance I needed, I had to include things such as a specialized 5x8 multiplication routine and a similar asymmetrical divide function.

 

Ultimately, I got into a showdown with my boss about underbidding jobs with impossible schedules. He decided to demote me and I left.

 

Basic Four                                Position/Title: Group leader, three projects, acting Manager of Software Engineering

1980-1982                                 Synopsis: (Z80 assy & PLZ; OS design),

                                                Details: Primary project responsibility was for the operating system for the Spectrum 80 (later shortened to S80) business computer. The operating system was a disk-based system using a B-tree file structure supporting applications written in the company’s proprietary Business Basic.

 

Tuboscope                                Position/Title: Independent contractor

1979-1983                                 Synopsis: Pipe inspection systems (6502 assy; microcomputer graphics).

                                                Details: Tuboscope wanted to build an inexpensive pipe inspection system with a graphical display. They had gone to Apple when it was just a startup and Steve Wozniak hand-coded (on a legal tablet) a prototype system using an.Apple II motherboard. The project got shelved for six months and when they were ready to start it back up, Apple had hit the big time and Wozniak was VP of engineering and no longer interested in proceeding. I was one of the only people in Houston doing contract and consulting work using 6502 µPs, so I inherited the job. I finished the design and additional coding, then took the project through three more revisions over the years.

 

Fann Instrument                        Position/Title: Electronics manager

1975-1978                                 Synopsis: Managed & developed rheological instruments. For last 8 months, acting director of R&D (CMOS logic, analog design; rheology).

                                                Details: Managed all electronic instrumentation activities, including engineering, R&D, and production. My job also included conducting training in instrument usage and basic rheology at various customer sites (typically in Oklahoma - e.g. Phillips in Bartlesville, Conoco in Ponca City, Amoco in Oklahoma City, etc. – or Louisiana). Specific Projects included:

·         Model 50D system viscometer. A precision rotational viscometer with shear rates up to 600 RPM and temperatures up to 500 °F. The shear rate was programmable with ramp up-ramp down over various time intervals. The temperature was similarly programmable. Both the motor speed control and temperature control were servo systems using a precision tachometer for the motor and a thermocouple for the temperature. Implementation was with discrete 4000-series CMOS logic and analog (op amp) circuitry.

·         Model 39C bench top viscometer. This was like a baby model 50 viscometer, with programmable shear rates, but no programmable temperature and pressure. The technology was similar. This and the 50D were my “from scratch” designs.

·         Model 23C emulsion stability tester. This used a high voltage probe to detect the voltage at which an invert (water in oil) emulsion would break. The high voltage section consisted of a blocking oscillator. A simple comparator detected the knee in the V-I curve. I redesigned the 23C primarily for cost reduction.

·         Model 88B resistivity meter. Before I left, I redesigned and built a bench prototype of what would eventually become the 88C.

·         The Consistometer, a high pressure (20 kPSI), high-temperature (500 °F) instrument for measuring relative viscosity (“consistency”) of  casing cements and other fluid using in extreme conditions. Although out of production now, the Consistometer lives on in the Fann model 75 HPHT viscometer, designed using principles I introduced before I left.

·         I also maintained all of Fann’s legacy products.

 

Western Geophysical                 Position/Title: Project, sustaining, design, and test engineering

1972-1975                                 Synopsis: Geophysical instrumentation (logic design & DSP)

                                                Details: As a sustaining/test engineer, I was responsible for  dedicated test systems which automatically tested the company’s ruggedized 800/1600 BPI data recorders and cost reduction of our dedicated autocorellator. As a junior project engineer, I was responsible for the design of the “Manpaq” analog subsystem – part of a field portable complete seismic system. The analog subsystem contained the preamps, filters, and 30 channel roll along switch. Other subsystems were the battery pack and digital recorder modules.

 

Other experience:

 

·         Published in Dr. Dobb's Journal, C User's Journal, and Windows Programming Journal.

·         In late 80's, created and have since maintained www.snippets.org web site of free educational resources

·         ANSI working committee for HPPI bus (ANSI X3.218-1993).

 

Education:

 

Rice University, physics.

 

Relevant memberships:

 

IEEE, ACM, APS, WINA, ISA, Mensa

 

Personality:

 

MBTI: ISTP; Enneagram: 1/5; Big 5: O93-C17-E3-A32-N22

 

References:

 

On request.