@ Supply Chain Management

Adrian Gonzales ruminates on his eleven year stint at Arc Advisory in the post: The Accidental Analyst: My 11 years at ARC. It’s an interesting read about the progress of an analyst. Adrian talks about what has changed and what hasn’t changed in this arena:

[The dotcom era]. era served as a catalyst for companies to “think differently” about how they could leverage Internet technologies to transform their supply chain and logistics processes.

and,

Data quality remains a huge problem in the industry, and it’s only getting worse. Garbage still flows in, and garbage still flows out.

Companies still want their 3PLs to be more proactive, and 3PLs still want their customers to view them as partners instead of suppliers.

Everyone talks about collaboration, but when times get tough, as it did last year, it’s every company for itself.

Forecasts are always wrong.

It’s about people, processes, and technologies, but not necessarily in that order.

Data quality is never going to improve because data means different things to different people and so also the classification of what is garbage and what is not. The Internet is one big garbage dump but that’s no reason not to dive in – there’s a pony in there somewhere. Look at it this way, we’re in the stage of global search aka Google, Bing and what have you. Next, will come local search – simply because people will become attuned to the notion that they understand data better than some bot running on a google server in Sunnyvale, CA. And so on.

Forecasts are always wrong just as such statements that imply the same. Forecasts are mostly wrong because it is about the future. I’ve talked about that on my blog several times – there are two extremas to making forecasts – very detailed ones and a few milestones here and there. In my opinion, what is critical is not forecasting and its accuracy but execution and its implied capability. I’d focus more on short run forecasting with rapidly adaptable capability than the converse. When the converse is the case, people lament about the fact that “Forecasts are always wrong.” Yes, but that need not trouble anyone nor detain them.

Let me leave you with a sanguine observation about bubbles. The last bubble cannot be reflated because of the knowledge everyone has about it, especially about its bursting and the fact that a majority of the people are still vested in the last bubble and have yet to realize that it is not coming back. That implies that the mania of housing is over for a good while and that asset class has to go back to being a stodgy preserver of wealth (whatever little there is of it).

The consequence is that without an asset class so commonly found and the method of extracting value from it as the asset grew to bubble proportions, consumption cannot be financed at the previous pace. That has many implications for those who create goods for consumption and those who supply it (and transport it).

So all hail to the next bubble – contrary to the notion that the masters of our fate are pulling the strings that bind us; it is only that we’re so poor string pullers that we end up stringing ourselves up nice and dandy.

An article summarizing the keynote speech of Glen Hodgson, senior vice-president, and chief economist at The Conference Board of Canada is available here: Supply chain management ‘critical to success’ going forward, says Conference Board of Canada’s chief economist.

From the article

Technically, he said, while economic growth has returned in the US, reaching 5.7 % in the last quarter, Hodgson said there has yet been no strong recovery in private investment and consumer consumption.

The US will pay a high price for the heavy government investment that has taken place to boost the economy, he said.

and,

“At the peak of the housing bust in the US, one third of mortgages were subprime, ‘NINJA’ mortgages, i.e. no-income, no-job mortgages. The one piece of the US financial workout that’s not yet been settled is the housing market. There’s no true stability,” said Hodgson.

While there is still a significant overstock of housing inventory in the US, the supply is down to six months, and housing starts are predicted to rise but not at a great rate, he said.

About a year ago, the supply of homes was closer to nine months and so it is good news that some of that excess has been worked off. However, that in itself doesn’t mean much because homes are not yet affordable relative to income. Furthermore, unemployment still persists.

Going forward, Glen Hodgson predicts

“US consumers make up 70% of US GDP, and 15 % of the worlds, so a drop in consumer confidence is significant. China will eventually take this over as Chinese consumers move from saving 42 % of their income and gain access to credit. Americans are actually saving again-about 5 % of their actual paycheck,” said Hodgson, who predicted a ‘temperate recovery and consumption’ for the US, which is now borrowing about 1.5 billion dollars annually, a hole that represents about 10 % of GDP.

“They will have to have a serious debate about taxation,” he said.

The US will grow at about 2.8 % this year but the balance of ‘too much government’ is wrong, and represents what Hodgson called a ‘ U-shaped recovery’. He said the US won’t experience sustainable growth before 2012.

So that’s the take on the current situation as well as general forecast going forward. But why is Supply Chain Management critical to success going forward? The reasons don’t change even if the circumstances may:

In five to fifteen years, the rise of India and China, an aging demographic within the industrial world, and integrative trade and global value chains will be additional factors to deal with.

Supply chain management becomes ‘critical to success’, said Hodgson, who noted that firms having an international strategy “built around integrative trade will be in better shape.”

This is precious little to chew on for supply chain practitioners. However, I think we have known about this shift that will take place which is a fundamental driver of economic growth – a consumption hungry population (and a huge one to boot) that is getting to binge on credit.

This has all the makings of a disaster and from the looks of it Toyota Corp has the deer in the headlights look. This has all the makings of a PR disaster minus the PR.

The driver of a Toyota Prius who called 911 on Monday to report his accelerator was stuck finally got the car stopped after about 20 minutes with the help of the California Highway Patrol, officers said.

"He was reaching speeds over 90 miles per hour," CHP Officer Larry Landeros said of the driver, James Sikes.
A Toyota spokesman said Monday evening that the company, which has recalled millions of vehicles because of reports of unintended acceleration, was sending a representative to investigate the cause of the incident.

Here’s the video of the story:

I think the situation is reaching the point of breaking and if Toyota designers cannot find the root cause quickly and/or implement a safe workaround until creating a final fix – It does look like Toyota’s reputation notwithstanding TPS is about to careen off a cliff.

But the company was unsure whether Sikes took his car into a Toyota dealer to comply with the recall, Lyons said.

Why? The driver may not have but the company is unsure?

What happens when Toyota falls flat? The price of success is hubris and there is for the hangers on a measure of schadenfraude to be had. To put it delicately, hot water is a kind description for the fabled auto manufacturer – there is even a website that I found that tracks issues with various Toyota models : Toyotoproblems.com

Logistics Management has an article on Toyota’s problems: Supply chain: Analysts weigh in on Toyota recall. The summary of that article is as follows:

.two analysts who spoke to SCMR today can confirm one thing: Toyota dropped the ball when it came to analyzing reports of problems in the past, reports that might have helped the company avoid such a dangerous and costly issue. Both analysts agreed that the chief lesson here for Toyota, and any other company looking to avoid similar problems, is to have tighter, more efficient analytics to better spot problems in advance.

and

Joe Barkai, practice director at marketing intelligence and advisory firm IDC, cited a 2008 report by the National Highway Transportation Authority, the government body that controls recalls in America. That report, he said, includes internal Toyota documents dating back to 2003, where the automaker detected an "unintended acceleration issue," which at the time had been attributed to floor mats which were not properly anchored.

This is clear proof, Barkai said, that Toyota has had problems for a long time with acceleration, but has failed to act.

"This recent Toyota incident is definitely not new," he said.

Michael Burkett, vice president of research at AMR Research, came to the same conclusion, citing the same documentation. There is clear evidence, he said, that Toyota was not processing the data from accidents and service reports properly.

"They were having difficulty correlating that information," he said. "It doesn’t appear that they detected problems were arising until it came to a head."

A key component in this fiasco seems to be the following that Joe Barkai points out:

Barkai suggested one factor in the size of the recall is the very design of Toyota vehicles, which in this case, may have been too efficient for the company’s own good.

"The cars are very, very similar," he said. "They use the same parts. They use the same suppliers."

As a result, Barkai said, even a minor parts-related issue could easily affect millions of cars at once. Of course, Barkai added, this same factor should have allowed Toyota to see the problem a lot sooner.

Taking the above observation along with the mistaken conclusion of the root cause analysis of “unintended acceleration issue” which was improperly anchored floor mats – the issue of the sticky accelerator pedal becomes a little clearer. Is that hubris, cutting corners or something else – some time in the near future, I think, the answer and solution will become clearer.

As you will observe, there is a feedback loop in place (falling back into Control theory terminology). What happens when the feedback loop input is not being correctly valued? The suggestion made in the article above is for better analytics in order to sift through the incoming data for yet another look. This is the role of the Observer in Control Theory. Then we get into the proper design of observers and their many inputs and the output – but it still comes down to resolving the feedback more than having an observer in the loop.

After a little more digging, I looked at some design related news about the same problem – sticky pedals. The article is from Design News: Toyota’s Problem Was Unforeseeable.

Toyota’s sticking gas pedal was an almost-unforeseeable problem, experts say, and the best course of action now is for engineers to ensure that drivers can handle the failure if it happens again.
"This is one of those horrifying nightmare problems that will occasionally occur, no matter how hard you try," said David Cole, chairman of the Center for Automotive Research.
Automotive experts said this week that predicting the problem would have been nearly impossible during design and test, especially given the kind of accelerated testing that is typically used to evaluate components which may have to last from 10 to 15 years. Making it even more difficult was the fact that the gas pedals didn’t appear to fail by themselves, but rather, by interaction with other components, such as heaters or floor mats.
"It’s not that they didn’t design a good accelerator pedal or linkage or floor mat or heater," said Steven D. Eppinger, professor of Management Science and Engineering Systems at Massachusetts Institute of Technology (MIT). "They designed them each quite well. But the most difficult problems always relate to interactions between components and other systems."
Although Toyota now appears to be coming close to a repair for the gas pedal problem, many questions still remain about its genesis. The giant automaker has gone through a succession of theories about the problem’s cause, including interaction with floor mats, materials in the accelerator’s friction lever, and condensation and corrosion from heaters. During the two-year course of problems, Toyota has examined its floor mats, shortened its pedals, lengthened the friction lever and changed its linkage materials. This morning, the company reportedly said it will add a "spacer" that will increase the tension in a spring that would keep the pedal from sticking.
Still, experts say that one of the best fixes is one that helps drivers deal with the problem when it happens. "The takeaway is that it’s less about durability testing and accelerated testing, and more about designing for failure," said Jake Fisher, senior automotive engineer for Consumer Reports.

The solution put forth there is to implement throttle-by-wire systems

Toyota’s throttle-by-wire systems, already in place on most or all of the affected vehicles, will soon contain additional software commands that will interrupt the flow of gasoline to the engine if a driver hits the brake pedal. Such software could go a long way toward preventing fatalities, since most drivers instinctively step on the brake pedal when the gas pedal sticks. Many competing automakers already incorporate those software commands in their electronic throttle bodies.

Whatever the root cause of the problem and the solution for it, there is little doubt in anyone’s mind that Toyota has stumbled and stumbled badly. What has been even more damning has been the sluggishness of the corporate response over and above the engineering, sourcing and manufacturing issues that may be at the heart of the problem. For that a heavy price must be paid.

SC Digest has a recent article: A decade in Supply Chain Timeline illustrating a decade in the supply chain timeline. What astonishes me about the decade is how quickly the mighty have fallen. Or rather how quickly some of the no-brainer trends seem to go quite limp.

The stars from the report in my opinion are:

1. Blue jeans icon Levi’s announces it is shuttering all its US production and moving to Asia – as do others throughout the decade.

If anything this move has probably been one of the most important signals of the decade.

2. Yellow Freight announces plans to buy Roadway express, its larger LTL rival. That plus other acquisitions soon bring financial trouble to new YRC Worldwide.

A telling example of how the golden boys gamble and fail.

3. Walmart announces plans for RFID tag mandate, later says it expects all vendor pallets and cases will be tagged within a few years

and later towards the end of 2009

Walmart’s RFID program seems to go into total limbo.

Now, that’s an example of how the go it alone mandate may not be the best way to introduce far reaching changes in the supply chain. Is that also a signal that the supply chain has exhausted ways of making things cheaper and in the next decade of inflation and rising commodity prices, these costs are going to find their way through the supply chain directly to the consumer.

4. Mattel becomes poster child for concerns about offshoring generally and product safety for goods made in China specifically as it has to recall millions of toys due to lead in paint and other issues.

An idea of the true costs of offshoring begins to emerge as unit cost is balanced by risk related costs and soaring transportation costs. No free lunch indeed.

5. Independent truckers going out of businesses by the thousands as slowing freight volumes and soaring fuel prices take their toll.

The driver shortage reported by American Trucking Association earlier in 2005 was most likely filled by independent truckers capitalizing on the shortfall. But the onset of the recession rearranges the marketplace again.

And the cake on the icing:

6. Dell pulls big surprise by announcing in April (2007) quarterly earnings call presentation that it is entering the retail market and largely abandoning its legendary build-to-order supply chain model, saying it is too costly.

If you survey the history of manufacturing management, you’d obtain a sort of cycle that comes and goes. The form of the cycle is centralization and de-centralization of the operations ostensibly for cost savings and better efficiencies and what not. Ergo, it behooves me to make the obvious prediction that Dell would go back to the make to order business before the end of this decade. But that is predicated on a recovery in the general economy.

7. Factory utilization reaches post-Depression low of 65% in the US.

After a decade of outsourcing and offshoring and, thus climbing up the value chain, we begin to appreciate that the perch up the value chain can only accommodate a few and the rest have to be paid for my debasing the medium of value i..e money. Fortunately and unfortunately, this game can only go so far.

And finally,

8. Everyone looks forward to 2010 – perhaps most difficult year to forecast in decades.

Au contraire mes amis, this is the easiest decade to forecast let alone year. This decade is going to be the lost decade – perhaps, I should wait at least until the first year is out. Maybe not.

Here’s wishing you and yours a Happy New Year ahead. It promises to be an interesting year indeed.

Cheers!!!

I’m writing a book and I’ve put up one of the chapters from that book aptly titled Supply Chain Heresies (for now). The chapter that I’ve completed is called Continuous Improvement. Download it and send me critiques if you’d be so kind and so inclined.

Have a great weekend.