Automation patterns…any figures?
Market Watch reports:
Hewlett-Packard Co. on Tuesday said it will invest $1 billion to consolidate and automate its data centers, a move that will lead to the elimination of 9,000 jobs over a “multiyear” period.
The Palo Alto, Calif.-based computer and printing giant /quotes/comstock/13*!hpq/quotes/nls/hpq (HPQ 45.59, +0.01, +0.02%) said annual savings in its enterprise-services division would range from $500 million to $700 million once the transformation is complete. H-P plans to halve the number of data centers its operates to around 50.
Data centers are used to store computerized information of businesses, government agencies and other organizations.
…
At the same time, H-P said it would hire more than 6,000 workers in sales and other areas of its enterprise-services unit.The shift toward automation will allow H-P to better deploy assets obtained via the acquisition of EDS in mid-2008, according to the company.
“We see the possibility of taking enterprise services to an entirely new level based on applying H-P industry-leading technology to the services business,” said Cathie Lesjak, H-P’s chief financial officer.
From the HP website:
Tier I: Proof-of-concept testing, porting services, Web testing, solution benchmarking, product/solution demonstrations and training, customer briefings.Tier II: Solution benchmarking, product/solution demonstrations and training, customer briefings.
Tier III: Product/Solution demonstrations and training, customer briefings.
Are Tier 1 and Tier 2 jobs becoming Tier 3 jobs for people? The trend in this sector has HP a bit behind competitors.
What is a data center in a company that makes PCs, one for each person in the company? HP has always confused me, eliminating a 1960s technology in 2001 to replace it with an 80s technology they supposedly invented.
Surely the data centers support their IT service business (aka “Enterprise solutions”), not the hardware manufacture and sales.
You have to have HP as a customer to appreciate what a mind boggling large company they really are. An income statement or press report just doesn’t do it justice.
Some terminology:
A call center is where a consumers’ call gets routed with a tech or customer service question. The physical endpoint for just about all Fortune 500 companies now is somewhere in India. (official language is English)
A data center is the IT place corporations have and it consists of network routers, servers, and biz software and applications. Some is generic and much of it is custom software developed for a specific company. The trend has been to outsource this monstrosity to specialized companies that know how to tame the beast.
In olden times, HP was only a manufacturer of tech products, but they have been pursuing an IBM-like biz model where consulting, software development and data center outsourcing are playing a much bigger role in the biz model.
DOD did the same thing all through the late 80s and 90s. What it does is dramatically reduce: redundant facilties, undersized facilities which can be combined, and facilities using old technologies. IIRC DOD went from well over 1,000 facilities to just 15. I don’t remember the savings, but it was in the 100s of millions.
I suspect it is mostly a consolidation of the EDS centers, and then a further consolidation into the HP centers.
Yes, EDS was bought and absorbed. I was unable to find out what kind of jobs were being eliminated at what $$, where the buld were, and compare to kinds of jobs being proposed, the $ for the new jobs, and where the bulk were being added. I don’t see how 3000 jobs equals 750 million savings. a year.
HP is doing this to be competitive in a competition where automation is ‘necessary’. It is even a good business decision. It is a notion of what replaces the jobs and where.
“consolidation” is the right word for it. Biz systems already are automated. The trend is to outsource the “physical plant” to companies that run huge data centers. This will reduce the headcount of network engineers, security specialists, and probably maintenance programmers (customers always want bugs fixed, and mods made to existing systems) because they will be more efficiently used.
There is also a trend for smaller consulting companies (biz system developers) to become experts in a particular industry and tweak custom systems they developed for one customer at huge expense and do a “IT modernization” project for another customer in the same industry which can use essentially the same system. Then they oftentimes will run and maintain the system either at the customer site, or at their own data center and the customer is hooked up by WAN or secure Internet.
It’s all to “save money”.
Last I checked, which has been a while ago, a senior programmer in the US made $80K as a company man with bennies, and if very lucky and very sharp, up to $120K as an independent contractor.
However in India, a PhD in computer science got $20K, senior BS got $10K, and junior code monkey got $5K.
The trend also has been to shift the work to India.
“The trend also has been to shift the work to India.”
I’ve heard this stated quite a few times, but have yet to see any actual data on it beyond anecdotes. Is there any actual evidence this is happening on a wide enough scale to matter?
Around year 2000 IBM announced that over the next ten years, half of IBM’s workforce will be outside the US. It didn’t all go to India of course, they are a multinational with biz everywhere.
But it began in the early nineties. The Indian government suggested to the population of India that they should become computer programmers. Sort of like Japan did with electronics and cars. Next thing we new there were about 4 very large Indian programming consulting firms competing for contracts in the US. To counter the thrust, all the major US consulting firms, IBM, CSC, EDS, etc…began setting up shop in India. The “business analysts” worked with customers here, then sent the actual system design specs to their shop in India to develop the code.
“We” were competitive once again.
Nothing happens all at once or 100%, but it was and is very significant. You can google “Bangalore” for info on the new city for computer programmers. Or listen to webcasts of the IT consulting firms and hear how it’s going straight from the CEOs mouth.
Of course Uncle Sam helped with handing out about half a million multi-year visas. You can hire them here too, from temporary employment agencies, then send them back to India after you have had your way with them. They then study up on more “technology” so as a single organism they will out learn you no matter how hard you try, and will have an edge up on that next contract over someone working with “old” technology.
But we had to get them trained up with experiance somehow.
There’s a good post that talks about the HP announcement and the impact of “cloud computing” here:
http://industry.bnet.com/technology/10008400/hp-layoffs-uh-service-improvement-%E2%80%93-are-cloud-computings-future/
I also have a post on my blog here:
http://econfuture.wordpress.com/2010/06/02/housing-as-an-atm-on-the-way-up-and-on-the-way-down-plus-automation-at-hp/
“The conventional wisdom, of course, is that if that happens, there will be jobs for all those workers created in other areas. I really wonder: Where, exactly?”
We are all going to be bankers, lawyers and politicians.
Almost forgot the military. Now there’s the place for robots, you would think.
Not real sure what they are talking about when the say “automation” as it applies to a data center. I chose leisure time over geek heaven in 2003, made a years severance pay buying CSC and ACS stock, and waited out the growing housing bubble in S. Cal. So I’m not real current with the state of the art in IT.
But network engineers, security specialists, database administrators and programmers all use “tools” to do their jobs, and they do keep making these “tools” work better and therefore increase our productivity and as a consequence reduce our numbers. So maybe that’s what they mean. But I would still worry more about Bangalore.
Anyway, I wish someone would hurry up and automate doctors. That would be the perfect fit for an “expert knowledge” system. No machine intelligence necessary, Just step by step diagnosis.
I think we definitely we see very powerful AI-based diagnostic tools pretty soon. The problem is that doctors are protected via licensing requirements, etc. Same thing with pilots, etc…even thought the technology may be there to automate, it is unlikely to happen anytime soon. That’s why I think knowledge workers sitting in cubicles are going to be highly vulnerable. They have no such protection.
I think most people would agree with your perception that Bangalore is more important that automation, but studies have shown it not to be true.
http://www.oecd.org/dataoecd/37/26/36187829.pdf
And there is evidence that offshore workers also get automated once the technology is there.
OK, I made it to page 11 and got tired of reading. The report does say they don’t have much data to go by. Now when they say it’s not a big percentage of the IT service sector, I guess I could swallow that statement, because over the period they attempted to study, the only IT job that could be outsourced was programming, and programmers are probably a minority compared to all types of jobs summed up in IT. But if you are a programmer I would go by the fact that Bangalore IS there, company webcasts say they are doing that, and our job website, DICE, doesn’t return ay job hits anymore (tho part of that reason was most companies stopped doing anything new starting late 2000. I got 20 a day before then)
Also, I didn’t see if they include call centers or not. Pink collar took a hit there.
But agreed Cloud Computing will make an even bigger implosion. And the combination of automation, and being more secure and stable service in the US than overseas perhaps, may be the thing that keeps it here. We’ve got to put something in all our office space.
***I don’t see how 3000 jobs equals 750 million savings. a year. ***
I’m guessing, but I think that the job losses are just a side affect of consolidation. If we assume a fully burdened cost of $120000 for each of those lost jobs (and I’d bet that is on the high side), that’d be 360M a year, the other 390M would come from projected non-employment related savings due to consolidation e.g. fewer backup generators to maintain, lower overall property taxes, etc. Like I said, Just a guess.
***I think we definitely we see very powerful AI-based diagnostic tools pretty soon.***
People have been predicting that for about 60 years and so far they have been consistently wrong. But I suppose that someday they’ll be right. Maybe this time it really is different.