In the previous post:
the queue theory links together telecommunication and supply chains management (SCM) since both actually share the same issue: how to deal with queues?
In order to achieve a certain Time To Delivery, according to queue theory, there are two approaches:
- increase inventories in order to compensate demand volatility.
- increase velocity that means achieving low lead times (production, supplies, deliveries) in terms of material flows and information flows as well.
Anyhow, how to face market demand volatility providing to customers the same Time To Delivery (velocity) without increasing inventories (waste of resources)?
Nowadays markets are more volatile than in the past and customers are expecting short time to delivery (see the same day of delivery battle between amazon and Ebay).
How to be more reactive toward changes in market demand?
Market Volatility, Frequency and Shannon-Hartley sampling theorem
A market demand is commonly drawn in a chart and it may looks like in the picture below where:
- T = is the reference period when measuring the market demand. Actually, with regards to S&OP (Sales & Operation Planning), is when customers’ orders are aggregated for defining the schedules for the next period. (es. 1 month).
- Si = is the level of the demand (aggregated orders) at time i. For example S1 = 200 pz in January (i=1), S2 = 150 pz (i=2) in February and so on.
Well, according to Shannon-Hartley theorem T is the sampling period. Intuitively, the shorter the period the better signal changes are tracked.
In particular, given T the sampling period and B (bandwidth) the highest frequency that is wanted to be recorded, we have that:
T must be lower than 1 / (2*B)
And what does this mean in practice when dealing with demand volatility?
As an example, it means that if customers’ orders are aggregated once per month, no more than a change within a period of 2 months can be managed by the company.
No matter if MRPs, TQM, Lean Manufacturing and other tools are in place.
Batches, Queues, One Piece Flow and Lean Thinking
According to Lean Thinking (see to James Womack), the challenge is switching from batch & queue paradigm to one piece flow manufacturing. That is the philosophy for reducing waste: inventories and time.
Batch & queue paradigm is the main cause of waste, both in terms of high inventories (low liquidity for investments) and time (effectiveness).
Companies that are re-thinking theirs business model towards a one piece flow approach will achieve operational excellence and market effectiveness.
In practice, achieving one piece flow means being able to operate with high frequencies(*):
- processing customers’ orders once a day instead of once a week
- production planning once a week instead of once a month
- production scheduling once a day instead of once per month
- and negotiate with suppliers weekly instead of monthly deliveries.
…and here where issues come out.
Attempting to speed-up processes and activities along the value chain (production, order management, deliveries, supplies,…) it’s actually an opportunity for:
- discovering unproductive tasks along the flow, both material and informational
- useless complexity (product mix, BOMs,…)
- technology legacy (machinery, IT systems)
- the need for company training and personal engagement on continuous improvements
There are all constraints for improving velocity and TTD.
(*) point 1,2 and 3 are part of the S&OP (Sales and Operating Planning)
…So: can you guess which is the “bandwidth” of your company? Ask to Mr. Shannon and Mr. Hartley
- markets are getting more and more volatile
- TTDs required by customers are getting shorter and shorter
why frequency and tack time are so important in lean manufacturing as well as the demand planning period for S&OP?
It can be inferred that, according to Shannon-Hartely sampling theorem, it’s because the demand planning period (S&OP) is strictly correlated with the ability of the company to react on changes in the market.
T is the planning period so that only one change in the demand within a period of 2T can be handled by the company without safety stocks.
From another perspective, what if the company attempts to reduce the planning period T?
There is a T_limit, due to many constraints such as lead times (suppliers, cycles,…), set-ups, shipments,… where a further reduction of the planning period will not produce significant improvements in terms of inventory reduction and TTD.
Again, thanks to Shannon-Hartely sampling theorem, the T_limit defines a kind of company bandwidth, a KPI for measuring the ability of the company to mange market volatility without increasing inventories.
T_limit is the shortest planning period where no significant improvements in terms of inventory reduction can be achieved by increasing the freguency planning, and the company bandwidth B is defined as 1 / (2*T_limit)
- company bandwidth B = 1 changes / 1 month. That means that the company can react to demand changes once per month without increasing inventories.
- the T_limit is equal to 1 / (2*B) = 1 / (2*1) = 0.5 month. Thus, lowering the demand planning period below 2 times per month does not produce any significant improvement and inventory reduction.
So, since Lean Manufacturing implies achieving one piece flow and ability to operate with high frequencies (bandwidth),…
can you guess which is the bandwidth in your company?
Feelink – Feel & Think approach for doing life!