process-improvement-solutions.com Blog

This incident is in the news again. The French authorities have resumed the search again for the black box. The Root Cause(s) of this incident were never conclusively established. A National Geographic indipendent investigation implied that when super cooled water came in contact with the pitot tubes (speed sensors) it froze all the sensors (four?) and the control system lost air speed thus deactivating the auto pilot. The pilots, due to error message overload and having lost track of the throttle position (which does not adjust to its actual setting as in older electro mechanical aircrafts), were not able to set the aircraft to safe thrust (85%) and wing angle causing a catastrohic stall.

Nat Geo’s investgators catogorically ruled out Lightening strikes as a possible cause (as alluded to by me). Not sure why the control system would not default autamatically to the safe thrust (85%) and wing angle when disengaging the auto pilot.

In today’s WSJ there is an article on Indiana’s problems with its Social Services. The claims process was outsourced to IBM and its subcontractor, Affiliated Computer Services. This is an excellent example of how automating a process without first understanding all the underlying process issues does not really fix anything. In fact it makes the problem worse as discussed in my Amazon.com review of the book published by the founder of Infosys.

Most IT projects fail because the process issues are not addressed first. It is a common fallacy to assume that simply installing technology systems and automating a process will resolve all issues. The problem in this case does not appear to be outsourcing the claims process, but rather, lack of effort upfront designing the right process to fix the issues. The constituents using the system are the customers here and really the entire process needs to be revamped from outside in – customers’ point of view.

Shaker Cherukuri
Managing Principal
Process Improvement Solutions, Inc.
Greenwood, IN
317-258-3552

There is a lot of debate going on about our Energy Problem and how the current stimulus is trying to tackle our Energy crisis. Yesterday, President Obama unvelied a $2.4 Billion grant to the private sector with the primary objective being to boost innovation in battery technology.

Solar firms in California have been significant receipts of grants in the energy renewables sector. This specific grant targeted towards battery technology in the automotive space was targeted towards firms in the midwest especially in Indiana and Michigan such as EnerDel (automotive lithium-ion batteries will receive $118.5 million in a matching grant), Delphi Automotive Systems LLC ($89.3 million), Allison Transmission ($62.8 million), Remy Inc. ($60.2 million) and Purdue University ($6.1 million).

Just as the Strategic Defense Initiative (SDI) program (aka Star Wars) funded research (mostly in the defense industry) in the 80s spawned numerous innovations even though the SDI as envisioned did not pan out, this (the current fiscal stimulus) too is likely to sow the seeds for innovative solutions for our gargantuan “Energy” problem.

The entire eco-system needs evaluation and new solutions – transportation and its power source (i.e engine/motor), fuel, fuel source, power generation, distribution, consumption. The current stimulus actually hits all of these aspects of the energy eco-system and is using the private sector and consumer incentives to change the behavior and investment/development focus.

Obviously not everything will bear fruit, even if 20 to 30% of all these efforts succeed, we would be well on our way to becoming an energy efficient society.

Shaker Cherukuri
Managing Principal
Process Improvement Solutions, Inc.

DC Metro Collision blog on YouTube
It appears that the rail cars in the metro collision in DC were owned by some banks and leased back to the DC metro authority to extract the value of the tax asset (depreciation). DC metro does not pay taxes and the banks do. Structuring a financial transaction to extract the maximum benefit from a capital asset is fairly common practice in most capital intensive businesses. For example, aircrafts are leased by most airlines since they usually do not have earnings and cannot benefit from depreciating the asset at the front end. So GE and AIG are the biggest lessor of aircrafts (AIG I think has spun off or created a separate entity for this business) since they get to depreciate the asset and the airlines get to be the lessee.

The life calculation of the asset and the associated risk analysis should be a critical component of such a financial transaction since essentially you are choosing to operate the asset for certain duration. The question in the DC metro case is: Whose responsibility was it to maintain the rail cars in operating condition? Was it the bank that owned the cars or was it the operating authority? It does not seem like that age of the cars was a contributing factor in the crash. However, newer rail cars most likely could have mitigate the impact of the collision.

That is true for any asset in general – Cars, homes, appliances etc. Newer models have more safety features and can protect the users in case of accidents. Does that mean we replace everything frequently? How often? Obviously that is not realistic. It is a tradeoff. In the case of airlines, the maintenance, repairs and replacement life cycles are the responsibility of the airlines or the suppliers like GE that sign a contract to maintain the engines or in some cases even own the engines and the airline pays by the hours of usage. So GE (capital) owns the aircraft, leases it to the airline and then GE Aviation owns the engines and the airlines pay GE aviation by the hour for the use of engine!! GE has a dedicated team that does all the risk assessment for these capital assets and the financial gains from these contracts are thoroughly audited. These teams are shared resources between GE aviation and GE Energy since the energy business enters into similar contracts with the power plants for the Gas Turbines that share the core technology with the aircraft engines (I was a Senior Program Manager at the GE energy risk assessment team while at GE).

I digress. Getting back to the issue at hand (DC metro collision). The collision was most likely caused by a procedural system failure. The control system failed? Maybe. Was it designed to detect the stalled train? Were the standard operating procedures followed by all involved? Did the standard procedures account for the chain of events that occurred? Was there an environmental factor? Did the control system malfunction? Was it the controller or the sensors or the software or a combination? Did the life of the rail cars and its components play a factor in causing the crash? It will take an exhaustive Root Cause Analysis to get the bottom of such an event. It takes time and a structured approach.

Shaker Cherukuri
Managing Principal
Process Improvement Solutions, Inc.

Air France Flight 447 Root Cause Analysis – A possible hypothesis < Youtube
It appears that a chain of events and the associated "system failure" caused this crash. A fully automated, highly sophisticated aircraft with multiple backups designed to make the system fault tolerant cannot simply fail. However, if one were to expose the system to an environment that is beyond the design limits and them combine that with human judgment error, then a very highly improbably cascade effect can cause the failure.

Electronic components in computer systems are susceptible to spurious noise signals induced via transducers that either malfunction or fail. To preclude this, there are usually multiple transducers (sensors) and multiple controllers in a multimillion dollar aircraft. It is hard to believe that all these redundancies failed at the same time in combination with human error followed by structural failure. An exhaustive Root Cause Analysis will most likely take a long time and lot of resources with multiple projects to get to the bottom of this unfortunate accident.

Shaker Cherukuri
Managing Principal
Process Improvement Solutions Inc

“Quality is a marathon, not a sprint” – That is the key statement made in yesterday’s WSJ article about GMs Chevy Malibu by the Toyota spokesman. What does he/she mean by it? Read on….

Sustainable quality is possible only by inculcating a culture that enables the creation and propagation of quality. All the processes of the “system” need to be self regulating and self correcting (the famous TPS principles). In the short term, it is possible to produce a quality product via extensive inspections and “sorting”. This increases the production costs by generating lots of scrap due to low yields.

The initial quality here is a reference to the infant mortality part of the bath tub curve. If you just cherry pick the good parts, you can reduce the infant mortality (i.e improve initial quality). The failure rate of the population during the useful life is a function of good design margins and tolerance limits. This can only be improved by better designs and manufacturing processes. The trick is to design the product right to meet (exceed) the consumer’s demands (VOC) and then create a manufacturing process that is capable of manufacturing all components, subsystems and systems consistently well within the tolerance limits so you do not have to have a quality inspection at all!! The durability is whole different issue. This is where things like galvanized steel used to make the BMW chasis (discussed in a earlier blog) help make the system more durable.

This sort of eco-system takes years to create since you have to partner with your suppliers and enable them to produce like you do. The Malibu is definitely a very good product. With a lot riding on it, GM put lot of effort into it. The question is, did it pay off? Is it profitable? Is the quality the result of an excellent design and manufacturing process or is it due to extensive inspections and sorting? Is this sustainable?

The bigger issue at GM is that there are too many products. If I want to buy a minivan from Honda or Toyota, there is only one option. Why does GM have so many variations of the same product?

Now is a great time for GM to differentiate and reposition itself by offering products and services that offer a unique value to its customers. The legacy fixed costs are being reduced due to external forces (which was not possible before). GM needs to capitalize on this gift and focus on creating the capability to run a marathon and not just a sprint.

Shaker Cherukuri
Principal
Process Improvement Solutions Inc.
3209 W. Smith Valley Road, Suite 224
Greenwood, IN 46142
317-258-3552

The basic premise of a March 9th WSJ article on Lean is that in an automated system, the managers are unable to cut costs by letting people go. There is no one left to let go. There is all this capital equipment that is collecting dust, but no one wants to let it go. It is sunk cost. Substituting labor for capital (equipment) when demand is high might be the solution to increase capacity, the reverse might be needed when demand is low as eluded to by Toyota in the Feb 24th WSJ article. The decision needs to be made at the margin and not be based on the sunk cost of the capital equipment. Lean is about matching your supply chain and internal capacity with demand – not just having less people (labor) and more machines (capital equipment).

Looking ahead, with less demand, it might be cheaper to go back to an older process that required more people and less equipment – which is what Toyota was taking about. As for temporary issues in a JIT system causing issues with the supply chain or internal operations – that is usually an acceptable tradeoff due to the benefit of a JIT system over the long term. Dell did not abandon its JIT supply chain from Taiwan when an earthquake in Taiwan disrupted its supply chain for a few weeks. The shift from 90% of people in agriculture to 5% today is structural. What I am talking about here is a tactical short term issue of substituting capital equipment for labor since the variable cost of operating the capital equipment at lower capacity might be higher than producing at lower capacity with labor which can be phase in incrementally which may not be possible with the capital equipment.
Shaker Cherukuri
Principal
Process Improvement Solutions Inc.
3209 W. Smith Valley Road, Suite 224
Greenwood, IN 46142
317-258-3552