Hope to continue with my mission. And, as always, comments/suggestions from readers are welcome.

1. Doxobid, a Dr. Reddy’s brand for 400mg doxophylline is reportedly sold at Rs. 83. While same formulation under Macleods generic brand is sold as Doxoril at Rs. 43.
2. An injection was charged for  Rs. 30, 950 by a retailer to hospital, while distributor’s/hospital price should’ve been Rs. 25,543. Additional cost was, of course, transferred to the patient who received that injection. There is a FDA investigation going in this case.

Alright, second example was not about the difference of prices among generics and brand drugs…. but nevertheless, such instances strengthen NPPA’s proposal to standardize drug pricing. With ~5% of any branded drug price spent on promotions, brand-owners will have very little margin left if such regulation gets approved.

Generic-makers are not in sweet spot either. If brand-makers are forced to sell some of their drugs at lower price, they have capability of flooding the market with this lower priced, but better “perceived” quality drugs. Market share struggles, in that case, will force generic-makers to get their margins higher to recover loss in market share.

In any case, both generic and brand makers need to reconsider their cost-structure now.

Another global player trying to introduce alternate fuel vehicle is GM. They are avoiding heavy import levies by partnering with Bangalore based Reva and developing an electric vehicle (EV) indigenously. By doing so GM-Reva are planning on launching affordable EV by next year as well. But, they don’t expect significant volume. Tata Motors is also planning to get Indica EV out in markets abroad a year ahead of it’s introduction in India.

One of the major limitation for owners of an EV, that would need to be plugged-in to get “fueled”, is power infrastructure. As per Indian government’s 11th 5-year planning commission (2007-2012), there is peak power deficit 15% in India. With 90,000 MW of capacity addition targeted for 2007-2012, they plan add twice the capacity planned during previous 5-year period. And, then there are challenges of transmission and distribution of this power.

So, auto-makers seem to ahead in the game right now. With right incentive and infrastructure in place and appropriate pricing, they should be able to churn out volumes with next couple of years. Indian customer is a smart one with an eye on overall economy of owning a vehicle. So, it won’t be a hard sell at right price, provided they have charging stations available throughout the country.

It’s good old, optimal production planning practices. A study by Ross Systems of pharmaceuticals in the US showed that improving forecasting and planning process of pharma manufacturing by 15% can lead to increased plant output of 10% and reduced inventory of 20%. This are some significant cost cutting means by which companies can improve their margins, or stay competitive at lower prices.

Better Forecasting

… is the key to plan for market demand. Market demand for drugs may be seasonal or maybe constant throughout the year. Having accurate historical sales data can lead to some good mathematical modeling of demand pattern and can give a basic model for forecasting market demand. This can then be combined with other, non-quantifiable market factors to prepare a demand forecast.

Demand Planning & MRP

For basics of demand planning and MRP you can refer blogs on this site. In general, this is long to medium range planning of demand and inventory. It takes into account 12-month rolling forecast to develop monthly/weekly production plan. Benefits of doing these steps right for pharmaceuticals would mean greater stability of plans, less changeovers, less reactive decision making, and lower inventory.

• safety stocks
• reorder points
• reorder batch size
• customer service level

… play important role in achieving this goal.

Typical output of demand planning and MRP cycle is a monthly and weekly production plan for each formulation at each of the processing facility.

Production Planning & Scheduling

The output of demand planning and MRP – a formulation/facility assignment – leads to short term planning at factory level. This is an intensive computational step. Information about current inventory levels, optimal batch sizes, batch processing time, relevant capacity, formulation contents and portions is used at this stage to prepare best production schedule that can…

• provide stable plans and reduce changes to production schedule
• reduce overtime by improving labour utilization
• fewer changeovers
• increased throughput
• reduced inventory level
• improved customer service
• improved morale and increased confidence in execution

No wonder then, that all these improvements can lead to significant cost-advantage for pharmaceuticals. With dwindling profit margins, pressure on costs will lead pharmaceuticals take this route in near future. Smart ones are already on their way of getting their whole supply chain under review.

Supporting Systems

Staying on top of current inventory levels would require real-time monitoring. There are many ERP solution available in market that can not only keep track of inventory, but also of processing data such as capacities, recipe, batch sizes, etc. Such systems can also be created from scratch and expanded on as-needed basis. For any system to enable best production planning for pharma business, it needs two essential components – an ERP system and an APS system.
ERP system keeps track of all the data input – sale history, inventory levels, recipe, factory capacity, etc.
APS (advanced production scheduling) system does all the computations based on adopted philosophy and goals programmed.

These two systems together make the magic pill that can help pharmaceuticals to stay healthy and battle-ready for oncoming pricing battle.

Two Factors for Seeking Advice

Knowledge and Criticality of the matter. These are the two factors I consider for deciding whether or not I need to seek advice on any matter.

As a general rule, if I have a knowledge about that matter (how to make my manufacturing more productive? – for example), then I don’t need a second opinion on that. But, knowledge by definition means “acquaintance with facts, truths, or principles, as from study or investigation”. So, someone else may have more knowledge on the matter based on their study and/or experience. For a critical matter, it makes sense to confirm my hypothesis by discussing it with someone more knowledgeable than me. The level of involvement of this person with more knowledge (call it an expert or a consultant) will be at different levels based on depth of knowledge I have about the matter.

After sharing my personal philosophy, I offered Mr. C couple of options to test my manufacturing excellence knowledge vis-a-vis his own. And, we decided to explore further for any critical challenges he may be facing.

Three Factors of Testing Knowledge

Of course, there is no standardized exam for testing somebody’s knowledge. So, options I offered to Mr. C included three factors that test depth of knowledge – brain, heart and guts.

• BRAIN – stores knowledge of the past. This is something most of the exams test… which is fine in academia, but misses two other factors of testing knowledge.
• HEART – shows personal attitude of the person. Every problem has (n+1) solutions… where, n = number of solutions one has already considered. Attitude plays a key role in picking up one of these solutions. Hence, it is important that while seeking advice from a more knowledgeable person, his/her attitude also matches or complements yours. Otherwise, you won’t be comfortable with the best advice provided to you.
• GUTS – indicates foresight and innovating thinking based on past knowledge and present attitude. No matter how much experience one may have, every problem is a new challenge and needs some degree of intuition to solve it. Building a case study and proposing a solution for that would demonstrate intuitive ability of your advisor.

Answering Mr. C’s simple question made me articulate the internal process of seeking advice that I’ve been practising for a while. Thank you Mr. C !!!

Customization of Process Improvement

Problems faced by companies are specific to their conditions. Even, companies in similar industry and during similar phases of their life cycle, they have their unique challenges. So, why use off-the-shelf tools and methodologies? Customizing process improvement programs (PIPs) gives companies chance to focus on specific problems, while still utilizing standard tools of improvement. There is nothing wrong with standard tools (Lean, Six Sigma, TQM, BPR, Kaizen, cost reduction), but it’s their reckless application that wastes lot of valuable time and resources. Hence, using such tools within a defined framework helps focus their effectiveness. Over the years, I have come to realize a refined way of approaching a PIP. This is an evolution of my earlier approaches where I used standard tools for PIP.

A Unique PIP Approach

Developing a customized PIP, that has long lasting impact and sustainability, is a process by itself. And, it begins with laying certain foundation. A right mindset goes a long way in leveraging efforts of all involved in this PIP. Training in certain hard and soft skills prepares the soil on which seeds of improvement can be sown with great success. Also, determination of right approach for improvement is key. On a broad scale, there are three types of management approaches – control management, incremental management and innovation/breakthrough management. Each of these styles is suitable for specific life-cycle phase in which company operates. The PIP discussed here is specific to incremental management style of operation. So, if your company is in the market of fading demand, then no incremental improvement will help keep it alive any longer. In such situation, there is a need to cannibalize your existing set of products and make entry into new markets. That’s innovation management and it needs entirely different mindset and skill-set.

After initial preparations, it’s top management’s job and responsibility to define goals for their organizations to march towards. Goals defined for such should have strong links towards that ultimate goal of making money (for business organizations, of course). Once that is defined, a complete chain of KPIs needs to defined that links highest level goals to smallest details of execution. For each of the KPIs thus defined, an accountability matrix can be defined to ensure functional division of labour to achieve the ultimate goal that is sought.

The next step is to manage constraints within each identified projects. In the world of diminishing resources, no team ever has time to address all the issues. Hence, concept of focussing on constraints is valuable from that perspective. For the identified constraints, scoring low hanging fruits is the best way to get snowball effect going. Tools of lean methodology are best suited for this. Demonstrating visible and early successes gets momentum going for next phase of PIP where, one would want to uncover hidden improvements with the help of initiatives from all involved. This is a stage where lot of data analysis may be needed to do cost-benefit analysis of any improvements. Six Sigma tools are best suited for this stage of PIP.

Most important aspect I learnt of PIP is following a “quick prototyping” approach. Whatever improvements are discovered, big or small, need to be immediately implemented. General FMEA is advisable to avoid catastrophic side-effects of improvements, but holding off such implementation till every small detail is planned and accounted for is waste of opportunity. Management support in this stage is crucial to understand philosophy behind rapid prototyping and not to punish small failures… but to focus on the pursuit for big victories.

Value Addition

• focus on INNOVATION - if business model of a startup revolves around offering creative solution, making sure that it really is unique and no other product is available in the market is of paramount importance. For manufacturing facility, this also includes providing unique manufacturing capability. But, even if the startup is not based on unique product/service idea, it must be unique in some ways. Otherwise, why would anybody buy anything from you? Innovation, is ideed, crux of entrepreneurship. 
• PROBLEMS are opportunities - the bigger the problem, the bigger the opportunities. Nobody pays anyone to get an email account. Or, to pays big bucks to get standard fasteners and sheet metal components. Businesses get paid for their unique ability to solve complex problems. 
• exceeding EXPECTATIONS - it’s not enough to meet expectations anymore. To stand out, one must exceed them… and even re-define them in the process. If the goal is to do just enough to get that next order of auto components, then one can get that order today. But, in this rapidly flattening world, what’s the guarantee that someone else will not come up with better quality, lead time and customer service to snatch that order away from you?

Inhibitions

• IMPROVISATION - there is no such thing as bullet-proof business plan. In fact many of the successful startups did not even have a business plan. Not to take away from importance of a business plan, but waiting for a perfect one is a wrong strategy. “Ready, Fire ! Aim.” is more appropriate strategy for entrepreneurs.
• fear of FAILURE - maybe number one reason why there are not enough startups. By nature, startups thrive on failures. Any venture capitalist can tell you that they don’t actually fund success, but they fund failures. Because, the percentage of failures is pretty constant over time. So, the more failures they fund, the closer they get to their next blockbuster. How is that good news for entrepreneurs? If they fail… fast and learn from them, then their chance of success greatly increases next time they try. Key is fail big, but cheaply (hint: use somebody else’s shoulder to fire).
• the LUCK factor - it’s not just by hard work that one can make luck; but there are lot other things by which one can make luck. As Tina Seelig points out in her talk, it’s not about making lemonade out of lemons… but it’s about turning lemonade into helicopters.

Business Side

• SCALABILITY - this has to do with vision. And, with rubber bands. A rubber band in your hand can be as long as you can stretch it. So, is any startup. If it is envisioned to be a big hit and scaled for such a growth, then it has a chance to succeed as it grows and realizes such a wildest dream of the entrepreneur. As my friend recently informed me, Twitters woes with Ruby on Rails is a recent and well known case of such scalability troubles.
• importance of CASH FLOW - it is hard fact of life. If you have few bucks to rent a private jet, you can pretend to be a millionaire who owns a private jet. But, if you don’t have a penny to pay for a cab… well you just can’t pretend to be anyone. In business, money talks louder than inspiring speeches. As an entrepreneur running your own show, one must either raise money or manage money or save money.

People Side

• TEAM character - this is a core team that will go through labour pains to give birth to the enterprise. Trust, understanding, passion, commitment, initiative are few of the character properties I can think of about a team that will make a successful startup… eventually.
• OWNERSHIP vs. MANAGEMENT - “letting go” is difficult for parents (entrepreneurs) when it comes to child (their startup). Yet, realizing that not everything can be done by one person (because lack of skills and/or time) is important. Would you rather compromise key functions within your startup, than getting it managed professionally by someone else? 
• getting BEST PEOPLE for the job - by “someone else” I meant…. someone capable. There is no replacement for a perfect (as perfect as you can get… remember, you can always improvise) person for a job.

To see complete discussion thread from which this blog is inspired, checkout LinkedIn group India Leadership Network and search for discussion “Why few startups succeed, majority fail?“.

There are two ways in which data can be collected from any tool – manual entry and thru machine interface. If the tool has some sort of interface capability, then using interface program (i/f) , data stream from the tool can be translated into useful information. This information then can be fed into main server (data warehouse). If such interface capability is not available, then manual tool time entry is the only option to collect some data. This would in the form of tool operator punching in few entries on a local terminal when tool changes over from one state (e.g. productive mode) to another (e.g. downtime). This will not be accurate as operator entry may be delayed or omitted at times. But, some data from tools is better than no data. If operators are encouraged to maintain accurate logs of tool time and are not harassed/punished for maintaining accurate data, then you’ll be amazed how zealous operators can get about contributing to this noble cause of watching your tool performance.

Data collected at local terminals can be relayed to data warehouse and can be stored there in a database. A report generator program on the server can, then, generate standard reports to communicate tool performance parameters to management. Depending on sophistication of the report generator, automated reports with charts can be published.

For manager of a modest machine shop, this may sound like a big undertaking. But, that’s just because I tried to impress the readers with whatever computer hardware/software terminology I knew. I talked with a friend of mine who does these kind of integration for living. And, it seems very doable on a modest scale. With availability of inexpensive hardware (used terminals, dumb-terminals, etc.) and open-source software, there is no need of reaching deep in one’s pockets to collect data from the tools. After all, these tools are most expensive resource items in many production facilities. If you’re running such an operation, wouldn’t you like to know how these tools are used?

Report generation is the information processing aspect of  this process and first three parts of this series will be of great help in deciding what KPIs to calculate and measure from the data collected from tools. Setting up such reporting can be counter-productive if KPIs are not balanced properly or linked to higher level goals. If you would like to try measuring your tool performance, then you’re already ahead in the efficiency game and I wish you good luck with your efforts in right direction. There are lot of topics I just touched upon briefly, which will play big role in setting up this system. So, use whatever production/industrial engineering expertise you have at your disposal.

Tool uptime indirectly shows how much time is spent in non-scheduled (NST), planned maintenance (MT) and repair time (RT). Using Lean terminology, all of these tool time categories are non-value added, as far as value to final customer is concerned. Hence, they should be eliminated. So, ideal value for any tool uptime metric would 100%. Period. There are qualifiers for this statement. Any tool engineering activity must be directed towards this end goal.

Next question for a keen observer would be…
Is, (10% of NST) = (10% of MT) = (10% of RT) ??
Mathematically, yes…. statistically, no.
NST is a non-scheduled time, for which facility is not available for production because of external conditions. These circumstances are related to either personnel issues (public holidays) or safety issues (dangerous weather conditions – rain, snow – that make coming to work unsafe for most workers). Though later cannot be planned, former can usually be planned well in advance. Such NST can be planned to create least amount of interruption to regular production flow. And, for unplanned NST (safety issues), one can always refer to the first edict of any production operations – Safety First.
Planned maintenance time, MT, is a preventative measure to carry out regular upkeep of the tool. The “planned” aspect of this time provides manufacturing enough warning about impending downtime and subjugate all their scheduling decisions around this.
Repair time, RT, is the most disruptive among all downtime events. It puts a brake to production activity when it is least expected and can result unbalanced production line. In some processes, it may result in quality issues in product batch during which this interruption has occurred. Random nature of RT can potentially create loss of capacity, and at the least it creates variability in production flow.

Tool utilization is an indication of efficiency in which tool uptime is converted into time during which product is created. One comment I got on Part II of tis series…. which was more of a question, was that should we not expect 100% utilization of every tool? It makes sense, right? Keep all the tools up for 100% of time and use them 100% of time they are available for production. Simple. This is what I also thought as I stepped into world of production engineering 15 years ago. Unfortunately, this is only possible in manufacturing utopia. Laws of diminishing returns and limited resources apply to tool uptime and tool utilization also. Beyond certain limit, every 1% increase in these tool metrics will see exponential cost increase. This cannot be more true than for utilization of tools in a production line comprising of several tools running at different processing speeds. In such scenario, there is always one tool performing certain production step, that is slowest among all. Overall production pace is dictated by this tool and is usually referred to in manufacturing as a “constraint” or “bottleneck”. This bottleneck is the one where you get most benefit on your resources. And, hence, this is where one must strive to get tool utilization of 100%. There is no real benefit of getting tool utilization that high on the fastest tool in the line.

Tool time, per se, is an indirect measurement of business performance. Linking it to monetary goals is very important. Typically, tools are considered as assets and are depreciated over time. So, depreciation value can effectively bridge gap between tool time and cost of production. There are other costs involved as well – energy consumption, consumables, service contracts, etc. Unit rate of such costs can lead to unit rate of tool time. How one calculates unit rate of tool time is very subjective. Depending on purpose of such allocation, setup time, idle time, engineering time may or may not be included in the calculation.

In final installation, Part IV, of this series, we’ll attempt to understand nuts and bolts of implementing such tool time measurement and calculation system in real life. Especially, when tools are not linked to any data gathering system, how do you measure this data and translate that into information.

• Tool Availability (TA%) – this is fraction of total tool time (TTT) for which tool is available for making product. As a consequence of this definition, non-scheduled time (NST), planned maintenance (MT) & repair times (RT) are excluded. This KPI indicates, on the scale of 0 to 100%, best-case scenario of how much of total tool time can be utilized for making products, even if tool downtime situation is not improved.
  TA% = (PT + IT + ET) / TTT 
   
• Tool Uptime (TUp%) – by definition of non-scheduled time (NST), there is very little that can be done to decrease this time. So, a KPI that helps creating actionable behavior would be one that excludes NST into tool availability calculations. Hence, another metric, TUp% can be used. This gives fraction of operational tool time (OTT) for which the tool is available for making products.
  TUp% =  (PT + IT + ET) / OTT  

  Like TA% above, TUp% indirectly indicates how much time is lost in tool downtime. Although, input in most management decisions, TUp% is really an indication of efficiency of maintenance group.
  A variation of this measure is productive tool uptime (PUp%), which excludes engineering time (ET) from above calculation. This way, net time available to production is revealed in a simple percentage calculation.
  PUp% = (PT + IT) / OTT
   

• Tool Utilization (TU%) – out of this uptime, only productive time is used to actually make products. So, it may be useful to track fraction of uptime (PT + IT + ET) for which the tool is actually working on products (PT). 
  TU% = PT / (PT + IT + ET)  

  One way to use this measure is to passively monitor it and plan accordingly. However, actively managed TU% can result in a much balanced line and best utilization of resources.
  Production specific version of this KPI is productive utilization (PU%), in which ET is excluded from above calculation. This gives exclusive insight into how well shop-floor supervisors are managing time available to them. It also, shows any inefficiencies in production systems due to lengthy setups.
  PU% =  PT / (PT + IT)

Discussion so far in Part I & II, provides framework of tool time measurement and resulting KPIs. Next, let’s see further interpretation and use of these KPIs and discussion on how basic data can be collected.