In mathematics, there are certain types of problems that are considered impossible to solve in a reasonable amount of time.  These problems are more common than you think and can appear deceptively simple.  For example, what the shortest distance between 85,900 particular cities?  That question took 136 CPU years.  These difficult problems are called NP-hard.  

When researchers attempt to solve problems, they need to determine if the problems are actually solvable, or if they are NP-hard.  Researchers know a new problem is NP-hard if they can transform the problem in question into an existing NP-hard problem.

So why did I bring this up?  I think a similar transformation might be possible between the IT infrastructure and industrial automation domains, but not in the way you might expect.

Industrial automation is about . . . automation

Recently I sat down with the guys at Automation.com to discuss trends they were seeing.  They pointed out that factories are starting to do extremely refined resource allocation to focus on the bottom line.  An example is a facility that balances (every few seconds) the cost of power and materials against the production plan to dynamically utilize resources in the most efficient manner possible.  The kicker?  Factories are doing this now using off-the-shelf software.

The IT transformation

When I heard about this incredible automation going on, it immediately sounded like an opportunity for the IT crowd.  As systems continue to migrate towards serviced based billing, costs will likely become more dynamic.  Let’s take the cloud as an example.

During peak hours:

-          your IaaS vendor may start charging a premium for disk writes

-          your PaaS vendor may start charging a premium for database access

-          your SaaS vendors may start charging a premium for throughput.

As an architect, you need to start thinking about all of these variable costs in your system design.  And you thought the cloud was going to simplify things!  So, how are you supposed to manage this incredible complexity?  Are you going to develop this expertise in house?

What would the pragmatic programmer do?

Reuse, reuse, reuse.  There is an unexplored opportunity to leverage the process automation knowledge and methodologies that exist right now.  It’s not just resource optimization – these guys also have lots of experience with real-time systems.  As we move to a dynamic world, reaction speed will become a critical part of any architecture and much harder to control as you inherit the performance of your weakest link.

I’m looking forward to further discussion with the IA community around the application of process control IP to the IT realm.  My hunch is that there are some killer applications just around the corner for the cloud vendors.

Am I nuts?  Time will tell.

Follow me on twitter @joshneland

I spotted this interesting commentary from Dan Herbert at ControlGlobal.com about using Industrial PCs instead of PLC or DCS solutions in industrial automation.  Dan puts together a reasonable take that highlights why using a hardened PC can save you lots of work and allow you to focus on the juicy stuff on your next project.

From the article:

Given this technology, there’s no longer any technical reason why a PLC or a DCS should be more reliable than an industrial PC. In fact, modern industrial PCs often use the same processors and embedded operating systems as their PLC or DCS counterparts.

Industrial PCs also offer many advantages in terms of cost and performance. High-end performance isn’t an issue when comparing PCs to DCSs, but cost and scalability are, particularly for smaller-scale applications.

“One cost-effective, but powerful controller can perform PLC functions for processes, run the HMI, perform high-speed data collection, and communicate to upper level computing systems,” explains Corey McAtee, product manager at Beckhoff Automation.

Advantages cited include

-          Suitable for harsh environments

-          Ability to support advanced and sophisticated features like HMI and remote access

-          Avoiding the cost of DCS in smaller deployments

-          Designed for longevity

-          Multi-core processors

-          Running multiple specialized OSes on the same box through virtualization

-          Avoiding the need for additional hardware when using a PLC

I have to admit, it’s getting pretty foggy out there for people who claim they need to make specialized computers when they don’t have extreme power or form-factor constraints.

Follow Josh on twitter @joshneland

This week I was wandering through the OEM engineering labs and stumbled onto one of our mechanical engineers, Tom Holt, who was putting away a sample of one of his latest custom bezel projects. Like a proud new father, Tom showed off his work and discussed with me the process it takes for an OEM customer to develop their own custom bezels for Dell’s PowerEdge servers.

Clearly Tom and his co-engineers can do amazing stuff for our customers. I cannot wait to see what they come up with in the near future.

If you are interested in learning more about how to customize a Dell server for your solutions, please contact us to learn more.

What happens when your software goes crazy?  Does your customer see it?  Do their customers see it?

To illustrate the importance of monitoring and management, I’m going to use a story.  This story is entirely fictional, but many aspects of it are entirely real for anyone who has a couple of product launches under their belt.

You choose the best and lose the rest

No one likes it when their products break, so architects typically design some fail safes into their systems.  These mechanisms monitor problems, and help in the diagnostic process.  If they are really good, the architect might even design the system repair itself when possible.

Then features start getting cut, because inevitably, your company only has so many development resources . . .  In this process, most companies end up forgoing manageability for a few pieces of shiny functionality, like a better splash screen, or more sophisticated SFDC integration.

So, your architect designed your product with a maniacal focus on the most important features.  Your team delivered on target with a few expected hiccups.  You’ve exceeded sales forecasts.  The future is bright.

A quality issue strikes

Just when you were planning your first project post-mortem, the support crew gets a call.  A unit in the field stopped working, and a customer is on the phone.  Your support staff and engineering teams huddle around a speaker phone, troubleshooting the issue remotely, trying to guide the non-technical customer through the process.

After a week has passed, you have made a plane trip to the defective unit and figured out it was a fan in the system that failed, and then the system ended up overheating.  You’ve had three more customer calls, but haven’t diagnosed them yet.

Your mind starts racing . . .

-          Will more units fail?

-          If so, when?

-          How will I know when they fail?

-          Do I need to wait for each customer to call me with a failed unit?

-          If I do a replacement campaign, how will I track which units have been fixed?

-          Maybe I have multiple issues?

Is there a better way?

Yes.

At this point, it would be nice if there was a way to interrogate the products in the field and take an inventory of their components.  It would be even better if the systems could notify your support crew each time a fan failed – and you could call your customer proactively.  You could even dispatch a replacement part anytime the thermals in the system began to increase – in anticipation of a fan failure.

Luckily, your supplier anticipated this event and built remote management of the hardware right into the box.  After discussing the issue with your supplier, you find out they supplied standards based interfaces to allow every piece of functionality you needed to get control of your products after they were installed at your customers’ facilities.

To top it off, your supplier even provided a management console that allowed you to do all of this right away from your home base.

After the product is launched, it keeps living

The best part about this story, is that because your supplier laid the foundation, you can continue to extend your products capabilities well after product launch.  These capabilities can continue to complement your current support pains and minimize customer downtime.

The moral of the story

If your supplier isn’t as kind as the one in this story, you need to take matters into your own hands.  Implementing the basic capabilities of monitoring, diagnostics and repair into your system can set the stage for delayed development.  In turn, that development can adjust as your product’s weaknesses become more apparent to you, and before they become apparent to your customers.

The 2^nd moral of the story

Make sure your supplier isn’t already fixing your problems before you invest your precious resources in duplicating them!

You can follow Josh Neland on twitter @joshneland

Three months ago my colleague, Josh Neland, brought to our attention a looming cost increase in RAM prices. At that time all of our customers were also being directly informed about this pending risk to prices so they could make arrangements to handle the change. Now we are seeing more references in the press to the risks to solution costs.

Hopefully every buyer out there was aware of this and has taken actions to prepare for the inevitable impact on cost. Planning is critical to getting through this and being prepared will make life easier for everyone in your organization.

Let me know how you are preparing. I’d love to know what works best.

I got the pleasure of visiting a new OEM customer last week when Corbin Moore and I delivered a cutting edge PCIe chassis from our DCS Lab to their lab for qualification. Corbin Moore, Business Development Manager for the Dell OEM Group, did most of the hard labor while I got to act as the cameraman. Fortunately, both the DCS labs and the NextIO labs were here in Austin and the drive in the hot, muggy 108 degree weather was relatively short.

NextIO creates and sells virtualized IO capabilities where storage and GPGPU performance is critical in HPCC and other performance cloud environments. Bob Shaw, application engineer for NextIO, shared the basic idea behind how they are using the new Dell DCS C410x PCIe chassis in the solutions they are developing and marketing.

This is a classic example of why I love my job. I get to visit some of the most technically advanced thinkers in the computing industry and learn about what the future holds for technology.

Last week our quality team helped speed up an OEM customer’s FDA requalification cycle.  How?  By opening up and sharing our extensive shock and vibe test data with a customer who was duplicating the testing in their own facilities.

The customer — let’s call them “MediDevice ”– has based its product on Dell’s OptiPlex XE platform.  Like many customers in the medical device market, MediDevice needed months to retest their product when a single component changed because of strict FDA guidelines.

MediDevice could have kept running their tests without asking if Dell had already run them, because they want to assume we just pump out boxes.  Instead, they engaged Dell at a deeper level and pushed on us with a real business problem.  The customer couldn’t change the FDA process, but they could change how much work MediDevice had to do after they received hardware from Dell.

Because their sales team (lead by George Frazier) listened and asked our product quality team what they could do to help, MediDevice found out that Dell was already running a good portion of their FDA required testing and certification as part of our everyday process.

The lesson – if you treat Dell or any other supplier like a box vendor, you will get a box and nothing more.  If you treat them like a partner and talk through your big issues, you might receive much more than you ask for.  In this case, MediDevice saved a month of testing on each new product revision just because they asked for it.

So go ahead and ask . . . what has your supplier done for you today?

My father was a cold war veteran. He worked for the U.S. government during the innovative period of the 1970s and 1980s when telecommunications first moved from mechanical to electronic switches and then made the tremendous leap from analog to digital. His involvement in the development of these technologies was reflected in his conversations at home with me and my brothers. Though I was too immature at the time to really appreciate his long and technical stories, I later realized what an amazing period that was for technology and risk taking. I now look up to my dad with as much respect and hero-worship any kid can have.

Dial-tone is essential

One of the core aspects of telecommunications I really took to heart was the absolute reliance any nation has on its ability to effectively communicate. This was the Cold War, after all. There were already multiple communication technologies available but the plain, old telephone wire was the absolute most critical and last line of hope should the unthinkable happen. So, the brilliant engineers at Bell Labs, working with the government and the sole telecommunications provider of the time, AT&T, developed the best technologies they could to ensure that the almost-magically important dial-tone was ever present. It would keep the nation alive and informed through any possible risk; be it storms, floods, earthquakes, or the omnipresent fear of nuclear winter. No matter what happens, there will always be a dial-tone.

Those genius technology thought leaders invented a strident specification called NEBS, which stands for Network Equipment Building System. They came up with three levels of the specification, Level-1 through Level-3, each adding more involved and increasingly-strict reliability, performance, and ruggedness requirements.

The NEBS specification proved to be very effective at ensuring the highest uptime for communications services. Even though more common communications and computing hardware now provides higher reliability, resilience, and redundancies, many communications companies still rely on the higher uptime promised by gear meeting NEBS requirements.

Cold war tech lives on

Today many communications service providers have moved most of their gear to more affordable, highly redundant commercial off the shelf systems (COTS). They still have deployments which require higher uptime assurances. In locations such as Central Offices which are scattered throughout urban areas to get the communications gear close to the end points where users plug in their devices, uptime features like DC power (to run directly from batteries), greater resilience to vibration (earthquakes), humidity, smoke and fire, etc. is still a common necessity.

The overall relative demand for NEBS gear has been shrinking since the end of the Cold War as the popularity of internet communications and cellular telephone use has expanded. But that reduction isn’t proof of the need completely disappearing.

What hath thou wrought?

I am curious to see how broad the demand for NEBS communications gear becomes.

What do you need from a NEBS solution? How important is long life when it comes to NEBS requirements? What are the levels of change management and locked down configurations gear?

by Josh Kivenko, OEM Global Marketing Manager

Staring at an “unusually uncertain” economic outlook in the US and deep spending cuts and write-downs in the Euro area (although industrial production data appears to be moving the in right direction) one would expect OEMs to show some short term price-driven thinking to how they are approaching their supplier relationships.

A recent survey by Riverwood solutions of senior OEM ops and supply chain professionals shows the opposite: only 14% of respondents (down 30% Y/Y) are looking at product cost as a top concern in measuring their electronics equipment partner and conversely, nearly 50% of respondents are willing to pay more for higher quality (up ~12% Y/Y).

While the aforementioned article focused more on supplier responsibility and its impacts to Chinese ODMs, I came away with a slightly different perspective on the data. The results confirmed what we’ve been hearing from our customers this past year: that value is important and that they are looking to gain a competitive edge by rationalizing cost across their operations. In other words, it does not begin and end with a procurement decision. Instead, OEMs are in search of PARTNERS that can help them control total cost and improve execution throughout their value chain. 

Case in point, we recently announced an agreement to supply Siemens Healthcare with a long life workstation to power their ACUSON SC2000™ volume imaging ultrasound system. What may sound as simple as a product decision was really a strategic one for this customer that was looking to streamline engineering, services and support functions by recognizing eventual downstream costs related to recertification and service strategy.

I also witnessed this first-hand with customers in Japan last week. I was in a meeting with at well known global electronics company looking for alternatives beyond their traditional contract manufacturing relationships. The customer was reviewing their next-gen design alternatives for a high-volume retail device, with an eye towards entry into the booming (yet completely foreign to them) Chinese market next year. They were grappling with the challenges related to their traditional custom design-oriented approach as well as the need to be more flexible in their operations and services model. We discussed how sensible trade-offs can be made by building x86 off-the-shelf technology into their design thereby adding a supply partner that can: help to them go-to-market fast, improve quality, enhance their support model and more easily extend their reach into new markets… all while maintaining their product brand and maximizing end-user experience.

In all, it looks like good news ahead for IT suppliers that offer a value chain orientation to meeting OEM customer need. Looking forward to seeing the full report from Riverwood soon.