Riffing on radical evolution in the ERP/MRP space

Profile picture for user gonzodaddy By Den Howlett May 20, 2015
Summary:
Brian Sommer's analysis about the future of ERP/MRP got me thinking about costing and supply chains.

3-D-printing-via-autodesk
Brian Sommer's blow by blow look into the future of manufacturing got me thinking about some of the points he made. I was especially taken with this section:
  • Seriously rethink what cost accounting and sales look like in the 3-D printer world. When parts have virtually zero labor costs included, virtually no scrap or rework, what is left for cost accountants to quantify or do? There may only be two costs that matters and that’s the raw material for the printer and electricity. The opportunity to be more efficient than a competitor may be lost. So, how will manufacturers compete in the future: cost (no), quality (no), service (possibly). In fact, great manufacturers will need capabilities to generate product as close to the point of need as possible, make it on-demand and deliver it right away. This reduction of the conversion cycle will impact sales forecasting, production scheduling and other functions immensely. Will the software evolve accordingly?

[my italicized emphasis added.]

Standard costing under pressure?

The costing issue is one that intrigues me. Even as a trainee accountant tasked to suck up arcane accounting concepts, I never understood the sense of standard costing. It's an arbitrary attempt at allocating overhead in situations where there is bulk repetitive manufacture for stock. It assumes that overhead is an important overall cost component needing attention. The problem with cost allocation is that it is incurred over time and almost always has no bearing on production units or productivity. That's why it is overhead.

The geniuses who dreamt up standard costing understood that but couldn't come up with anything better and in doing so spawned the whole notion of variance analysis which in turn seeks to explain how cost is over or under absorbed. If your head is spinning already then you're not alone.

I much preferred the marginal cost approach to understanding how manufacturing works but then I have a different mindset to the 'usual' cost and management accountant who was schooled in time and materials.

In Brian's imagined future, overhead takes on an entirely new meaning. I'd posit that it all but disappears because what might once have been considered overhead now becomes part of production cost. If you're hefting assets like IP as your unit production cost then it most certainly is the case. And the computing systems that support that management will be an integral part of what gets costed. Even more so if, as Brian suggests, the systems of tomorrow move to an opex consumption model.

If that's true then the entire accounting edifice gets massively disrupted. And not before time IMO. In its wake, a whole slew of compute configuration vanishes, putting paid to the many meetings that are usually needed to figure out how to make standard costing work.

But there is another problem that Brian skates over and which is far harder to solve.

Supply chain disruption?

Supply chains are an under discussed topic. Once a major talking point at a time when i2 was positioned as a leader in constraint based supply chain modeling, the topic has all but disappeared from the main technology discourse. This is an omission that gets resurrected big time in Brian's analysis.

A 3-D printer driven manufacturing scenario drives all sorts of fresh possibilities about how supply chains are organized. Do manufacturers of complex products for example go back to full vertical integration as was the case in Ford's early days? Tesla is leading the way in some senses but even they outsource important parts. If 3-D print technology allows for near instant production, then why would anyone want anything but a vertically integrated production plant? The cost implications of managing the supply chain would predicate against distributed manufacture in a world where instant gratification takes on fresh meaning.

If you buy that argument then some interesting scenarios arise. Which parts e.g. engines or gearboxes for example would Honda make for other manufacturers? Would they simply license their designs back to the main car assembler for onsite manufacture and effectively become an IP distributor? How would that work? Would Harley Davidson continue with its policy of putting bespoke finishing of its motorcycles into the hands of the retailer or drive down delivery by taking back in-house?

What about all those small but essential components made by mom and pop shops? They are often the part of the supply chain that remains the most brittle? Do these businesses end up being disintermediated? Could this mean the complete overhaul of plant scheduling? We recently saw interesting ideas playing out in chocolate manufacture in the Detroit area but sense these are scratching the surface of the possible in JIT delivery.

Oh - I forgot to mention the impact on logistics in a robot truck driven world of which more later.

My take

I've posed a few questions here as a way of kickstarting the kinds of conversation Brian envisages as the starting point for re-imagining manufacture. They are inevitably speculative but certainly worth considering.

We are told that modern software is only limited by our imaginations. ERP/MRP could well be the next bastion of tradition that gets blown up by what is available now. The only question for affected industries will lie in the extent to which they can re-imagine what 4th Generation manufacture looks like.

Gearing up to offer killer service will be the key to getting this right. How that plays out given the current state of the art is anything but certain but I believe we're not far from a point where real-time will be much more than right-time. It will mean punch in the online generated spec today, pick up the goodies tomorrow. It will be expected.

Image credit - Autodesk.