The Revolution in an Industry That Will Change The World (Test & Measurement)

Post on 19-Aug-2014

2.210 views 3 download

Tags:

description

The coming revolution in the $4B Test & Measurement industry that will change the world.

Transcript of The Revolution in an Industry That Will Change The World (Test & Measurement)

of an industry Revolution the

that will change the World

Test & Measurement

change the world?!?

we are obviously

CRAZY . . . right?

YZ

how could an industry you likely never heard of possibly

change the world??

stick with us, we will explain

✔  what “Test & Measurement” is

✔  why you should give a damn

✔  how a revolution in a historically boring industry will bring fascinating change

a long time ago someone built

something O O complex

and it didn’t work as expected

they needed to figure out what was wrong

swapping parts and guessing just wasn’t cutting it

along came someone with an instrument that could “test” and

“measure” the pieces and parts of that complex something

the right tool made troubleshooting tons easier and

saved a lot of

time

the concept of Test & Measurement was born

(more on the history of Test & Measurement as an industry in a bit)

why should anyone care about Test &

Measurement?

$113 billion identifying & correcting product defects

in the US,

is spent every year

let’s put $113 billion into context ~ ~

with $113 billion, you could ~ ~

end world hunger ($30B)

and buy

$30B $10B $4B $4B $4B

$4B $3.3B $2.5B $2B $2B

and still have $17B left

we sense your skepticism.

$113B is an unbelievably big number! what exactly is that being spent on?

the average engineer spends 1,500 hours / year fixing product bugs!! ~ ~

that’s 75% of our work time!

that means 75% of your engineering team salary goes to fixing product bugs!

oh wait, you thought we spent most of our time designing cool stuff?

wrong!! there are 100 bugs per 1,000 lines of

development code.

engineers spend most of their time finding and fixing these bugs!

$ time is expensive during

product development

$113B add up 75% of all of the product

engineering salaries across the US plus the cost of all of the debug tools they use

this does not even include the $$$ you lose when your product launches late b/c

of those pesky bugs

$113B+

and this is why Test & Measurement matters – a $4B industry (today) directly affecting $113B+

$

a history lesson

the invention of the integrated circuit in the

1950s set the stage for the electronics revolution

thousands of small parts joined together by invisible voltages,

currents, and connections

how do you troubleshoot a complex circuit if you can’t “see” the things that make it work??

back in 1946, C. Howard Vollum and Melvin

“Jack” Murdock invented the world’s first time

based triggered oscilloscope

an oscilloscope uses probes to capture voltages in a circuit and turn them into

waveforms that you can “see”

big problem meet

big solution

by the mid-1960s, Tektronix was booming as it dominated the o-scope market and Test & Measurement became a high-

growth industry*

* in the 1940s and 1950s, Test & Measurement was a much smaller industry primarily focused on tools for testing communications, audio, and video equipment

the 1970s welcomed the explosion of digital systems, microprocessors, and

inexpensive memory

o-scopes are great at capturing a lot of detail for a short period of time (i.e. they are good for debugging “analog” systems)

but they suck at debugging digital data ~ ~ (digital debug requires capturing a LOT of 1s and 0s over time)

in 1973, Hewlett-Packard introduced the logic analyzer, designed for digital system debugging

probe and capture large amounts of digital data, program trigger events that initiate capture, view digital

data as waveforms

for 30 years, the Test & Measurement industry was dominated by

oscilloscopes and logic analyzers

for 30 years, the probe was the product engineer’s debug tool of choice

in the early 2000s, Test & Measurement began

to completely change . . . and no one was talking about it

(especially those companies who had dominated the market for 30 years)

Marc Andreessen once predicted that “software will eat the world”

started to eat software

hardware

the rise of powerful system-on-chips (SoCs)

fall of memory and silicon prices

the 2000s brought

the

it was practical for us to put powerful processors and lots of memory inside our hardware products

we made our simple products substantially more complex, like giving them a brain

these “smart” products could execute lots of

complex code fast

software/firmware bugs

became the biggest obstacle in

the # of

product development

o-scopes and analyzers grew dusty b/c they were not built for software/

firmware debug

what tools were we using for debug?

embedded processor debuggers were the tool of choice for firmware developers b/c they could stop

the processor and step through code

but what did non-firmware engineers use for debug?

how did we debug if we did not have physical access to the main circuit board to

plug in one of these debuggers?

how did we debug in-the-field?

the rise of the log file

a log file is like a journal of what happened inside

a product

if software can detect it, software can “log” it

log files give us the ability to “see” what

software is doing

log files were a big hit

every product has a free, built-in “debugger” Z

log files are super-easy & super-fast to use Z

just retrieve it and read it

if given the choice between using an o-scope/analyzer or a log file, what

wins 99% of the time?

the easier and faster tool

this is the part where we let you in on a little secret about engineers

we don’t enjoy using o-scopes and analyzers - they take too long to setup

and the usability sucks

we don’t like sticking metal probes inside of products that are running

blowing up a $150,000 prototype because you stuck a probe in the

wrong place changes you

product development has become more software-centric

& the # of bugs to fix before product launch is exploding

today,

o-scopes & analyzers are no longer the center of the Test & Measurement universe

Test & Measurement Today (2014)

log files used by all types of engineers a lot

o-scopes / analyzers used by electrical engineers only when necessary

embedded processor debuggers used by coders when practical

there are big problems with the current approach to debug !

software (log files) can’t “see” everything we need

for product debug

there are times we need to capture raw hardware signals at a high rate of speed, something log files can’t do

for example,

what if software tells a pin to turn “on” but that pin is grounded “off” due to a hardware defect?

a software log can’t see this problem!!

there are times we want to use log files

but need to use o-scopes/analyzers

what tool do we use to debug the issues where software and hardware intersect?

hardware software

signals events

it’s a hardware problem

it’s a software problem

these problems cost an unreal amount of

time and $$$

these problems cost an unreal amount of

the revolutions are coming

we need tools that give us instant access to

anything going on inside our products ~ ~

~~

we need tools that are super-easy and

super-fast to use ~~

~~

we need tools that help us quickly understand

the data we capture ~ ~

we need the power of scopes & analyzers and

the efficiency of log files ~ ~ ~ ~

the power of scopes & analyzers and the efficiency of log files

we need

~ ~ built inside every product made

~ ~ ~ ~

the embedded instrument revolution

Moore’s Law has driven the size of transistors in ICs smaller and

smaller and smaller . . .

silicon real estate has become so cheap, it is common to find

unused silicon inside today’s ICs*

* a pad-limited IC cannot be shrunk any further because there would not be enough room for the bonding pads used for I/O pins on the outer perimeter of the

die, leaving unused silicon areas within the die itself

embedded instrumentation is the concept of inserting the

capabilities of external test equipment inside ICs

embedded instruments are controlled by a laptop, mobile

device, etc.

the power and functionality of your o-scope / analyzer shrunk down and placed inside the IC that was already going into your product

~~

for a growing # of today’s ICs, the

additional cost for adding an embedded

instrument is

$0

it gets even better!!

embedded instruments can do things scopes and analyzers could never do

(no, not this)

b/c they are inside the same IC that contains the microprocessor, embedded

instruments can capture hardware signals and software events, at the same time

hardware software

signals events

your scope, analyzer, and software logger rolled into one

~~ except, there are no probes

you won’t need them

the connected revolution

from 10 billion web connected devices today, to 212 billion connected devices by 2020

the ip address will replace the probe https://192.168.0.1

instantly access and troubleshoot web-connected products ~ ~

from anywhere

the software revolution

the future of debug is in the cloud

web software will collect data from web-connected products

smart platforms will know what type of data was collected

and deliver powerful, customized visualizations

log visualization-as-a-service

BIG DATA billions of web-connected devices producing trillions of data points

instantly available at the click of a button

information answers insights

why this matters

new product development will happen significantly faster

~~

better tools will remove a barrier to technology

advancements

service industries will completely change

imagine your next generation repairman remotely logging into your appliance and telling

you what is about to break before it breaks ~ ~

the financial and global impact will be profound

~~

the solutions to today’s Test & Measurement problems are going to open

doors of possibilities that will disrupt industries, spawn new markets, and

accelerate technology advancements

https://www.initialstate.com tools built for engineers, by engineers

Sources: http://www.ibm.com/developerworks/rational/library/4995.html Watts S. Humphrey, A Discipline for Software Engineering, Addison Wesley, 1996 Philip B. Crosby, Quality Is Free. Penguin, 1980 http://sqgne.org/presentations/2011-12/Jones-Sep-2011.pdf http://www.crosstalkonline.org/storage/issue-archives/2005/200504/200504-Jones.pdf http://inventionmachine.com/the-Invention-Machine-Blog/bid/87840/What-Are-Late-New-Product-Launches-Costing-You http://www.bls.gov http://oakstonepartners.com