Agile Manufacturing - Part 2

This blog post is in continuation to the first part of the AGILE MANUFACTURING blog series, wherein, we discussed how my thirst for applying Agile principles to manufacturing landed me at a dream project at an amazing manufacturing firm, the reasons for this blog series and the way I have organized/broken down my learnings at the manufacturing firm into various blog posts...

Let's continue from where we left:

Part 2: SMALL, INDEPENDENT, MEASUREABLE, TESTABLE tasks/stories actually help get (much) more stuff done...

As Scrum Masters, as coach and as practitioners of Scrum and other Agile ways of working, we keep promoting, practicing, advocating, suggesting, even shouting about how important it is to define our stories in a way that they are small, independent, testable and measurable (we even come up with hyped lingos like S.M.A.R.T.?and I.N.V.E.S.T.).

This simple practice went a long way in making remarkable changes:

CASE STUDY I - L.E.D. lights for the Indian Railways

Indian Railways is one customer with long-term warranty coverage, wide range of input voltages, strict quality checks yet inflexible deadlines with penalty clauses! This organization I have been working with, has been at the leading edge of innovating new LED based lighting solutions for railways to replace the existing CFL/traditional lamps thereby saving the running as well as maintenance costs for the country.?One challenge being faced by the team was painfully long wait periods due to long, inter-dependent chains of components and assembly:

While the design team finalized the outer box and worked with external teams for fabrication, powder coating etc, the optics team waited to fit-in the lenses perfectly , the lighting team in turn waited to fit the correct lumen LEDs on the lenses, the electronics team waited to work their magic on the "magic box" that converts any kind of electric voltage thrown at it (AC volt, DC volt, low volt, high volt, spikes, fluctuations...) into a smooth, uninterrupted supply of DC current to power the entire unit.

Clearly the team was that of experts and specialists who worked very hard but not smart. A clear solution that was missed out was to break down the process into SMALL, INDEPENDENT, MEASURABLE and most importantly TESTABLE units of work. This required mocking/stubbing out the dependencies, and THAT was the biggest show-stopper - not technically but psychologically - as they say - all problems are "people's problems".?We worked with the teams to understand their interdependencies, fears and possible failures in case we stub-out the dependencies. We then carefully incorporated and addressed all the listed risks and fears, while designing the stubs -

• The optics team only needed to know the cut-out specifics of the outer box, to be able to proceed independently
• The lighting team only needed to know the specifics of the lenses and the distance between them.
• The electronics team needed to know the series-parallel combination and the volt-current requirements of the LEDs to be able to proceed independently.

Once the list was complete, we prepared the stubs accordingly with all the testable parameters - voltage, current, wattage, spike, power factor, THD, lumens, dimensions etc clearly displayed with HIGH precision, digital display meters (did I say INFORMATION HIGHLIGHTERS )?- this was important to gain the confidence of the team. The management was initially not sure if we are heading in the right direction with money, time and effort going into building the stubs in a project "already squeezed for resources", but they went with my decision nonetheless. Pretty soon, the management as well as I were happy that they did - the turn-around time was magically reduced, the teams worked in parallel and independent of each other, there was renewed joy n energy and most importantly, the product(s) were shipped way before time.

CASE STUDY II - Product change-over turnaround time

Since this organization makes many variations of the products as per the customer needs, it calls for change-over of heavy die-moulds and related machinery, pre-heating of raw material for many hours, warming up the machinery, the chilling plant et al and finally resume the production.

The entire "change over" project used to consume a full day, and hence there was some resistance at frequent change overs owing to the overhead of time, effort, electricity, labor, down-time and so on. From a broad perspective it seemed logical as well.

However, when observed from a closer angle, it smelled of another "waterfall" running there... On careful analysis and breakdown of the units of work being performed by the team "in a sequential, natural" order, it turned out that many "parts of the work" needed many hours of waiting "after" they have been mechanically completed (for example, once the raw material is loaded, it is set for many hours of pre-heating) - Breaking down the entire chain-of-steps into smaller units of INDEPENDENT work, and re-arranging the same?with priority to those tasks that have a waiting period post-completion, did the trick. Now, the tasks with "post processing?waiting period" were done first, and thereafter other tasks were executed while waiting for the first-set of tasks to complete their "post processing". By the time, the "post processing" was done, rest everything was ready as well.

As a result, the team was able to reduce the "change-over" time from over a day to less than half and hence switching over to other production items/experiments was easier as well as costed way lesser.

CASE STUDY III - Die mould manufacturing in an "Agile" way!!

With its diverse portfolio of the industries that it serves - lighting industry, thinner and alcohol industry, farming industry, cooling industry et al, this organization has a very varied range of tools, machines and die moulds. Various moulding techniques like injection, blow, stretch blow, vertical injection, extrusion, co-extrusion etc are deployed - thus the die-moulds play a crucial part in the success of any product.

In India, however, the die mold manufacturing has largely been an unorganised sector. The time, effort, quality and price estimates are generally way off! Many innovations die-out simply for the want of the correct die moulds despite precise designs and specifications.

Also, there's an unsaid rule to not tinker with the expensive die-moulds, not because of the costs but the time involved to get them - hence a renowned die-maker had the "shock of his life" when he saw an expensive, intricate and high precession mould - lying dissected on our table - Instead of going with the traditional approach of making the entire die as the new one, we converted the die into a series of independent modules and changed just the module needed to be changed. The result of such breakdown into simpler, independent modules was this - the costs came down 10 - 50 folds (you read that right) and the turnaround time from more than an year to as much as 2 days - flat!! Not just that, this lower turnaround time opened a realm of possibilities for this "always ready to innovate and experiment" organization - that went on to experiment even more, leading to designing some "World's first and only" designs and products.

Anyone still wondering if these Agile principles work in the manufacturing world??

Next, let's see how - properly made and real-time updated INFORMATION HIGHLIGHERS make a world of difference - in this next post in the series...

<< Part 1

Part 3>>