Feeling Philosophical about CAD...

Where is CAD going? I know that when I'm in the retirement home, the kids
out of college are going to be fitted with Matrix-like jacks in the back of
their necks that allow them to directly create virtual designs simply by
thinking about them. What I'm curious about is what you all think we might
see in the next twenty years.

What will modeling be like? I suspect that we won't be working with
virtual lumps of clay, because I.D. folks don't do their concepts and
prototypes in clay in the real world - so will we all be shaping virtual
blocks of foam, or will we be using some other paradigm?

How will we be creating drawings? After all, we will still need them for
legal purposes, even if all of our vendors get online
..
What sort of input device will we use (until Microsoft develops that direct
shunt into the brain)?

What am I too tired (or dim) to ask that would be relevant?

I ask this after spending 30 Hrs over a beautiful weekend slugging it out
with a model that I've already got a hundred+ hours in. There are a lot of
really smart, connected folks on this forum - I am hoping that some of you
will chime in with an open-ended preview of what's to come, to give me a
little hope that, in the next decade, I won't be pissing away gorgeous
weekends struggling with software

Thanks for your thoughts/insights
-Ed

 

Remember in the sixties the way the year 2000 was supposed to be?
--> I'd be very much conservative.

The closest to come is, IMO, the 3D screen. Or real BIG Oleds.

Pointing? There are already systems that follow the movement of your eyes.
....Voice control?

Paper? I find much easier to find a document in my PC than any paper that I
have not used for a couple weeks. But paper is easier to read that screen.
Maybe "electronic paper".

....

 

Funny to see how you design guys focus on hardware aspects such as user
interfaces...
To me, Matrix and CAD are mostly software! Today's "CAD" with "D" for
"Design" is so close to the old "D" for "Drawing" that you need to add a
"3D" in front to make it clear that you're doing in fact CAGeometry...
Extrapolating from the last 25 years (see
http://cgw.pennnet.com/Articles/Art...Articles&Subsection=Display&ARTICLE_ID=134265),
I'd say CAD will evolve towards higher level of abstraction required by the
design of functional products, as well as integration of the various
technologies involved.
IMHO, future tools will use a formal description of the product
specifications, high-level functional modelling languages such as
www.modelica.org then MCAD, ECAD and other technology specific tools for the
detail design of each technologcial aspect of the whole product,
implementing specifications+manufacturing+cost constraints the same way as
geometric constraints are in SW now : red if what you do is too expensive,
red if it cannot be manufactured, red if it will break, everything wrong
will be red, and Ed will hold a dozen patents on the appropriate diagnostic
& error handling tools ;-)
So basically I say the future design will be top-down, and it will produce
documentation, virtual prototypes (software simulators) and CAM programs.
And you'll need an expensive add-on to output a drawing on the last printer
in your company.

Philippe Guglielmetti - www.dynabits.com

 

3d CAD is really cool.
But seductive.
After a day diddling and fiddling and dicking around with errant mates and
desktop crashes, I wonder just how productive I really am, compared with
early drawing board to Autocad (or in my case FastCad) transition which was
a huge productivity improvement.

As a matter of interest, I believe Douglas Aircraft Company had a prototype
DC3 flying 8 months after first pencil to paper.
They did this in St Louis at a time when the only air conditioning was an
open window and engineers and draughtsmen all had to wear neckties.

Hmm
malcolm

 

I wrote an article for Workstation News (long gone) in 1991 along the lines
you mention, and then I got a grant for developing an immersive system from
NASA and found out that you cannot tolerate HMDs for more than a few hours
at a time, if that, due to physical discomfort, low resolution, narrow field
of view and fear of head transmitted diseases. Women particularly disliked
them because they mussed their hair do. No way could you wear one all day,
all week, all year in a design environment.

Ten years on, HMDs are few and far between, and the resolution still sucks.
Gloves have a similar attribute. I think an HDTV screen with 1980 x 1024
pixels with a size of 42" diagonal in stereoscopic vision is all I would
need for design on my desk, based on Sony's new projector chip, but then,
for immersion, I'd use my flostation, the outcome of the NASA grant.
(pictures on request :>)

bp
flogiston corp.

 

Ok, I admit it, I'm an engineer that watches Star Trek reruns, big
surprise, huh? The first thing that you notice about any computer
interface in any of those future sci-fi shows is that there is no mouse.
I think that's a given for computer interface evolution, the display and
the interface will likely merge, unless the display goes holographic for
mechanical design. Ever since the typewriter, people want to see what
their hands are doing, and putting the keyboard and mouse on the display
will allow that.

As for CAD, I think that detail part design will to some extent be a
commodity generated by a knowledge based system. We used to be really
worried about the way our lettering on 2D pencil drawings looked. Now
that's something we get for free and we're instead worried about how the
parametrics in our parts will react to a change in the assembly. The
same thing might happen in the future when we change the design of an
assembly line that puts together automotive headlights and change it to
assemble cold fusion powerplants for vehicles.

Already I run into a lot of people interested in design automation.
Many companies design products that are similar, but have some
customization, or different sizes or options. It doesn't make sense to
hire people to design the same parts using different dimensions over and
over again. I see mechanical design becoming a function of setting up
the original design automation. People who actually do detailed piece
part design will be relegated to the "artisan" category. It will still
be a necessary function, but like strict 2D "detailers" will start to
fade into the past.

Companies won't buy CAD systems, they will buy a customized system that
allows website users to configure a product and give them a delivery
date, and the knowledge based system does the rest of the work,
including queueing the NC data to the manufacturing system.

I think CAM systems will change a lot too. Material removal
manufacturing is incredibly inefficient (duck, the machinists union is
gonna get me now!). The advances in additive systems (fused deposition
type processes) have been tremendous and may someday become ready for
high volume production.


matt

 

Hey, you could always fix that issue with a Holo-Deck as depicted in
Star Trek: Next Generation. No Head Mounted Displays, no 3D-glasses, no
gloves... Well, that my be more than 10 years away though.

 

Just as long as Microsoft isn't programming the operating system for
the brain/computer interface!!! Can you imagine your brain crashing
for no apparent reason in the middle of a critical application!

Cheers,

Rob

 

man you guys are no fun. there I was indulging in some escapism-flying
around like Superman punching louvers in walls ,welding with my
eyes.....followed by my loyal and shapely avatar : )......breaking off to
engage a Stormtrooper attack with an army of Bob Zees....and now I know my
future is going to be even more tormented by programmers with aspirations
toward expert systems....just terrible....

 

oh my...
bob z. is speechless...

--
bob z.
p.s.

"people with less brain power than you are doing more difficult things
everyday"©

 

Remember the old Steve Martin routine? ....Nah!

The reason you spend your weekends doing this stuff is not really because
the software screws you over, it's because you're crazy. Just like me and
half or more of the other folks who read this news group. If you weren't
crazy, you would either have planned on the software screwing up and
included the extra time in your schedule, told the client things were going
to be delayed a bit, or gotten another job by now. If the software didn't
screw up, you would find something in the design that wasn't quite exactly
right and fix it. Fixing it would take roughly 30 hours over a beautiful
weekend.

You're doomed. It's your fate. It's your weird.

Jerry Steiger
Tripod Data Systems

 

You're right - it is my fate.
Toget seriosu for a second - I can't bear to tell a client I'm going to be a
little late, because I know that every day that keeps their product from
market is another day that they lose profits from selling that product.
Enough days go by, and they have lost the salary of some guy that now needs
to be cut. I don't mind putting in a long weekend to keep some other dudes
family out of trouble.

And its not that the software screwed up - it ran pretty well (astonishingly
well, now that I look back on it. SolidWorks is kicking serious ass lately,
at least for me. Major changes that should have taken forever went really
smoothly). My beef, as I worked this weekend, was that I became aware that
its just terribly over-complicated. My forearm is sore from mouse clicks,
a third of which are obviously avoidable when you think about it. I'd see
hours tick away because of the process involved in making a change. If I
was on the board or using other software, I wouldn't even be done yet - but
that still isn't good enough. I want to know what CAD will look like when it
responds as fast as I can create?

I know you meant to be kidding around a little, Jerry, and I did take your
post that way (thanks) But it pointed out a good issue - we do not work in
a vacuum. Everything we design keeps tens or hundreds of other people in
jobs. I want to know what ideas you have about the day when it gets so fast
that we rocket through problems, so we can move onto the next project that
will keep another hundred folks employed!

 

Found this on a new type of keyboard

http://www.techtv.com/products/consumerelectronics/story/0,23008,3360537,00.html

 

As for CAD, I think that detail part design will to some extent be a

In my understanding , what you are talking about, is a case scenario
where if you had , lets say as a design problem, the top cover of a
flip-open type of a cell phone(thats the first thing infront of me now
) , you would just define the overal abstract shape, and things
such as locator protrusions, insets for the LCD Screen , fillets etc
would be automaticlly created.

Possible and cost-effective for a system ( or a KBE Tool if you may)
for a higly specific design problem with iterative/incremental design
releases. But intractable for a generic one.

It would yet require a system that accepts "knowledge" as input,
interprets it and commits the resultant computed geometric
manifestation to the design.

For a "from the scratch" design, The holy grail perhaps would then be
a complete and sufficient mapping of domain specific functions to
function shapes / shape creation techniques along with domain specific
relationship constraints between these function shapes -- This state
being extant for multiple domains.

We are also talking formalizations and knowledge encoding : which
means the Knowledge Engineer would have to go to work creating a
formalization and then a KBE tool for the specific domain. So if you
had a rough template geometry and wanted automated detail design on
this geometry, you would still have to specify functions at some level
of granularity, and assume that the system would commit detail design
features to the rough template automatically, know its location
automatically and its dimensions deduced automatically, resolve the
topological inconsistencies that may occur automatically ( a
self-healing geometric feature introduction) and decide to make it
consistent from a DFx Stand-point as well!
I wonder if thats really on the horizon when one tries to look at it
from a generic system point of view. There are a lot of CSP Problems
that are intractable. Or we might get stochastic and try genetic
design evolution.
But still, thats good late night reading.

Ucal Berkley has a pilot Agent Based Web-based System called
Cybercut.
which seems to be close.

Wouldn't be long before the only employees of FORD would be members of
the ford family with 5 buttons, one holo deck , a huge sales team ,
3000 disgruntled employees, and one EXTREMELY happy Design Automation
Analyst in a bomb-proof car.

Anup

 

A bright blue flame of rocket exhaust shot out of the back of the
engine, which sailed noiselessly through space on its way to the outer
planets. But Kyle, standing three meters away resting his chin on his
fist, knew the engine would be going nowhere unless they could
eliminate oscillations in the combustion chamber.

"I'm going to take off the cooling jacket," he called out to the
room as he reached for the floating holographic display panel on his
right and touched a translucent orange button. The cover disappeared
from the life-sized, holographic engine floating in the middle of the
room. Without looking at his instruments, Kyle could see that the flow
wasn't as steady as it should be.

"Let's take it one layer deeper and look inside the combustor."
Another touch on the floating panel split the engine in half,
revealing red fuel rushing into the combustion chamber forward of the
exhaust nozzle. The fuel ignited to blue in the combustion chamber and
shot out the nozzle into the vacuum of holographic interplanetary
space. But Kyle knew something wasn't right.

With the touch of another button, small yellow arrows,
highlighting small fuel concentrations that formed droplets just past
the fuel injectors, popped up inside the combustion chamber. That
wasn't the problem. The droplet formation was normal, but the sweeping
red ribbon-like waves bouncing from one combustion chamber wall to the
other were not. The acoustic waves meant there was a resonance coming
from somewhere. The waves grew in intensity until reaching a crescendo
as they entered the rocket nozzle. Worst of all, at the peak intensity
the hologram motion froze and the simulation automatically reset,
meaning a real-life rocket would have just exploded. The engine looked
like it was bleeding, and Kyle couldn't figure out how to stop it.

"I think we should call Rachel in," suggested Jeff, a structural
engineer with a soft Texas drawl. Startled, Kyle looked to his left.
Jeff was a little late, and Kyle hadn't expected him to be standing
over his shoulder. Jeff's image shimmered slightly as he moved closer
to the engine because he was actually standing in an identical lab
1,500 kilometers away.

Kyle muttered to himself and clenched his jaw. He was the lead
thermodynamics engineer on duty, and this problem was a showstopper.
Rachel's specialty was combustion dynamics, and Kyle knew if anyone
could solve the problem, she was the one. "Make the call."





--------------------------------------------------------------------------------




As the sun dropped towards the horizon, it threw a blanket of
gold across the lazy blue-green ocean. Rachel burrowed her toes into
the sand, concentrating on the grains rubbing against her skin. She
wanted to make sure she was awake, that this perfect sunset was real.
A fantastic end to an incredible day.

"Mom, you're being paged," Amber called from under an umbrella by
a grove of palm trees where she and her father had been sampling
drinks in coconut shells. Amber had finished her drink and was moving
her arms in slow circles in the air. Wearing her mother's
teleimmersion sunglasses, the young girl's odd ballet was really
movements in a gaming environment with friends scattered around the
planet.

Rachel took one last look at the sunset, turned, and headed up
the beach.

"Looks like they're having a little trouble," her husband,
Conner, said.

"It shouldn't take but a few minutes, and I am on call," she
replied.

"Come on Amber," Conner said getting up from his chair, "let's go
for a swim." Amber passed the glasses to her mother and headed toward
the water.

"I'll join you soon," Rachel added with a smile. She settled into
the beach chair and put the glasses on. A transparent heads-up display
overlaid the sunset as the voice of her computer-generated digital
assistant announced "Retinal identity verified. Welcome, Rachel. I
understand you have been paged. Kyle seems to be having some trouble
in the shop. I'll connect you now."

"Thank you, Rhett." The display immediately became opaque and
blocked out the beach and her family in the surf.

Almost instantly she was sitting virtually in the Immersed
Technology Design Facility, or ITDF in official jargon. But for the
people who thought of it as a second home, it was simply the "Design
Shop," or even just the "Shop." Rachel saw the holographic rendering
of the interstellar engine hovering in mid-flight at one end of the
room next to Kyle and Jeff.

"Okay, I'm here," Rachel called out. Kyle turned around and saw a
standardized image of Rachel. Because she was using a portable unit,
the system projected a constructed image based on stored data instead
of a real-time image of the beach setting.

"Thanks for getting here so quickly," Kyle said. "As you can see
we have a little problem." A swarm of red waves still poured out of
the engine.

"The waves represent acoustic wave fronts," Rhett informed her.

"So where are the acoustic waves coming from?" she asked.

"I haven't been able to figure that out yet," Kyle answered. He
was not surprised by the question. Although he could not hear Rhett's
comments to Rachel, he knew that the digital assistant would inform
her of the basic setup.

"Let's see if we can walk through this together and see what's
happening." Using controls activated by eye motion, Rachel moved her
viewpoint closer to the engine. Although she never left her beach
chair, Kyle and Jeff saw her image walk across the room. To get a
sense of what the system was doing, Rachel watched the engine run for
a minute. Kyle stepped up next to her.

"We've been looking at it for a while now with this same display,
but I can't determine the origin of the waves," he said quietly.

"I think I see something," Rachel said, as she moved a little
closer. "Set the controls to 75 percent flow rate, and adjust the
gimbal angle by, say, five degrees."

"Will do," Kyle replied, as he walked to his floating control
panel and started entering data.

Rachel couldn't help marveling that she was vacationing in
paradise with her family and, at the same time, helping to design a
rocket engine a continent away. Thanks to the flawless integration of
models and simulations, incredible advances in interactive and
communications technologies, and a cultural revolution, the
relationship between engineers and their work had changed drastically.
Here they were, designing the entire engine at one time, Rachel
thought. No longer were subsystems designed separately, to be
integrated later. System problems, such as this one, showed up
immediately. Design and testing had become one with advanced
physics-based simulation capabilities.

In the beginning there were a lot of challenges. Working in an
immersed environment was intense, and people tended to burn out after
about four hours. Maintaining a high level of performance day after
day, week after week, seemed impossible. But in the end they worked it
out with the help of an unlikely acquaintance. One of the design
engineers had a cousin who worked a rotating schedule on oil rigs in
the Gulf of Mexico. After a lot of argument, senior management decided
to try something similar, and it had worked far better than anyone
expected. The designers had been organized into two teams. Each team
worked for three weeks and then had two weeks off. Workdays included
four hours of integrated design and four hours of individual work in
their own offices. Some hours were devoted to debriefings and
brainstorming to improve the integrated design sessions. The change
had required a colossal shift in workplace thinking, but no one could
argue with the results. Time to completion was down and quality was
up.

Rachel was grateful for the change. She was spending more time
with her family and had a new sense of excitement about her work. She
was now on her two-week break, taking the vacation that had been
postponed several times because of the rough work schedules under the
old system.

"Okay, I'm ready," Kyle said. Rachel snapped back to attention.
Her display turned blue with red lines streaking past as she directed
her avatar to walk directly into the holographic combustion chamber.
Ignoring the red acoustic waves, she moved forward until she saw the
yellow arrows marking the fuel droplets forming just after the fuel
was atomized by the fuel injectors. The injectors acted like an
industrial-scale perfume sprayer, but with a punch. In this case, the
droplets were highly flammable rocket fuel that ignited almost
immediately in the combustion chamber.

"Kyle, could you please slow the flow rate." The fuel flow rate
slowed so that she could see where the fuel droplets formed. "Wait a
minute!" The whole design shop heard a loud pop when Rachel snapped
her fingers on the beach. "I think I have it, Kyle!"

"Where?" Kyle passed through part of her holographic projection
as he ran into the display.

"Right here. See how the droplets form all at once as they come
out of the fuel injector? See the same pattern that forms at equal
intervals?" Before Kyle could answer she said, "Rhett, what is the
time difference between individual droplet formation within a single
set of droplets?"

"The simulations indicate that 95.6 percent of the droplets form
within 0.01 milliseconds of each other. And the time interval between
droplet sets is a constant 0.12 milliseconds," Rhett replied.

"That has to be it. See the droplet formation here, Kyle?" She
pointed into the fuel stream. "It has a constant formation rate."

"Oh, that's it," Kyle sighed. "I should have thought of it.
That's well within the structural resonance range."

With new energy, Kyle turned around. "Jeff, please compare the
frequency of the mechanical resonance in the combustion chamber with
the frequency of the droplet set formation rate."

Jeff keyed a few buttons on his design panel. "They're within
half a Hertz of each other."

"That's it, then!" Rachel said.

"What's it?" Jeff asked, not understanding the consequences of
the data he had pulled up.

"The fuel droplet formation rate corresponds exactly to one of
the combustion chamber's acoustic resonance frequencies," Kyle
answered. "When the fuel droplets vaporize and burn all at once, the
periodic forces from the resulting pressure waves interact with the
mechanical properties of the combustion chamber. That unstable burning
process increases the peak pressure of the waves until the chamber
ruptures and the engine explodes."

"So all we need to do is shift the resonance frequency of the
combustion chamber," Rachel concluded.

"I'll take care of that," Jeff said, as he touched his display
screen a few times. "You two are a couple of smart designers."

As he finished speaking the red acoustic wave markers
disappeared. The combustion instability was gone. Everyone in the room
applauded. Images and voices of people who had been monitoring the
design work from other labs filled the room. Rachel had almost
forgotten about the other design and manufacturing groups that were
linked into the shop and had been monitoring their progress remotely.
One of the newcomers, a manufacturing engineer, called out to Jeff,
"What changes did you make?"

Jeff raised his voice above the excited noise. "First, I had the
atomization model change the hole distribution pattern with the
manufacturing tolerances in the fuel injectors to randomize the
droplet formation. Then I adjusted the dimensions of the chamber
slightly based on the chamber acoustic model to shift the chamber
resonance frequency. But when I adjusted the chamber size, the
structural model indicated I had to change the weave on the composite
fiber structure to compensate for some stress points the model picked
up."

"Well, according to the manufacturing stress models," the
manufacturing engineer said, "that composite change makes the
combustion chamber stronger. We can get rid of some supporting struts,
which also simplifies the assembly process. Good job!"

A moment later several team members jumped as an office cubicle,
complete with pictures of the family and a coffee cup pencil holder,
appeared at the back of the room. The cost engineer sitting at his
desk was flushed.

"Jeff! That's fantastic!" he yelled. "I just got the update from
manufacturing. Because of the reduced support structure, we can save 2
percent on engine manufacturing costs."

The whole room applauded again. Kyle cut in, "Hey, Jeff. Kudos to
you and Rachel."

"You're welcome. Call any time," Rachel smiled even though they
couldn't see her facial expressions. "But right now the waves are
calling. I'll see you next week." The engine they had almost finished
would be finalized in three weeks after a total development time of
five weeks. The preliminary parts would be available even sooner. The
rest of the spacecraft was being developed simultaneously in other
immersed design shops, and the mission was expected to take off,
literally, just nine months later. The schedule from initial concept
to launch was 14 months.

Rachel sat for a moment longer observing the excitement in the
room. Everyone was smiling and people were trying to slap each other's
holograms on the back and give translucent high fives. It was a good
feeling to turn a design problem into a design improvement.

"Disconnect please, Rhett," Rachel said. With a satisfied smile
she headed down to join her family for a sunset swim.

--- This actually is an extract from a report....and a driver and an
implementation for my thesis as well .....just
philosophizing....seriously

-Anup

 

sorry that was an oblique reference to Bob z. too....nice guy....my humour
probably doesn't travel well.

 

nothing more gratifying than spending some time in holographic goggles while
good ol' bob z. performs...

bob z. was there in spirit!
:~)>

 

Anup, that was the best part of bob z.'s day. really gives a fella
something to look forward to in 10 to 20 years.
:~)>
--
bob z.
p.s.

"people with less brain power than you are doing more difficult things
everyday"©

 

http://www.eetimes.com/story/OEG20030707S0066
closer than what we may realize.

--
bob z.
p.s.

"people with less brain power than you are doing more difficult things
everyday"©