Making Kinetic Warehouse Fulfilment Work in the Real World
How targeted Epicor Kinetic customisations closed the gaps between Advanced Fulfilment and the realities of a busy warehouse
Challenge
A move to Kinetic Advanced Fulfilment and mobile devices left niggles among the new efficiencies.
Outcome
After targeted work with users, errors and friction were removed, maximising the benefits of the new processes.
A manufacturer and distributer had moved to increasingly bigger premises as they grew, and reached the point where paper stock picking was unmanageable. Kinetic Mobile Warehouse was the answer, together with Advanced Fulfilment, Material Queues and Fulfilment Rules.
As a set, these modules and capabilities are extremely powerful, and transformed their ability to deal with a huge stockholding and rapid changing requirements from it, for dispatch and manufacturing material.
The problem
As so often, though, fitting real business processes into standard systems uncovered pain points, inefficiencies and trip-points for error in among all the massive gains.
Particular examples were:
Small touchscreens led to "fat finger" entries, with extra zeros or missing numbers leading to discrepancies between physical action and system record, all the way to invoicing the customers for wrong amounts.
Automatic location selection made picking easy, but it could be ignored without changing, so the stock and lot number could be different to what was recorded for the transaction, building up discrepancies.
Some items needed picking from bulk or high level stock, in which case a single material queue request couldn't cover the whole process, so fudged workarounds were needed for handing over between teams in a single recorded transaction.
Put-away locations were less consistent in reality than in system rules, which staff knew but the rules couldn't deal with.
PCIDs were introduced to control bulk storage, but breakdown of them had inconsistencies.
So overall, alongside big improvements in efficiency and throughput, errors and friction had been introduced, some of which noticeably affected customer satisfaction and required rectification.
The approach
This is not the kind of project where META eight, or anyone else, can walk in with a package that will solve the problem. The solution involved talking through each individual issue with knowledgeable staff, working with the expert functional consultants who had set up the processes, and creating small targeted customisations without affecting the core system.
Since most of the user interaction happens on mobile screens, scope for UI changes was limited. But through careful BPM creation we were able to add "are you sure?" pop-up messages for known signs of input error, adding checks for likely problems without friction when everything is normal, and require barcode-scanned confirmation of picking locations, with verification.
According to provided rules, we intercepted particular problematic Material Queue requests and tailored them, in certain cases adding new requests as partial completion was recognised, and in others altering the requirements in more subtle ways than the built-in systems.
We automated the splitting-down of PCIDs according to the organisational pattern, eliminating the errors and mis-steps from requiring it to be done manually, intercepting a UI flaw that made certain actions look as though they'd been done when the underlying reality was different.
In all cases, the customisations were constructed so that any failure would be soft, with warnings for the user, but falling back to standard underlying system behaviour rather than stopping the processes. In this way they could be confident that their worst case was always what they'd already accepted as default, and our refinements were safe.
The outcome
The immediate impact was an almost-complete elimination of the data errors, and close to 100% accuracy in customer delivery performance. This in itself immediately provided sufficient RoI for the work to be considered worthwhile.
Meanwhile the already large efficiency increases from the fulfilment procedures were further boosted by aligning what was happening on the ground with what the system required and recorded, reducing time spent on workarounds and increasing visibility. Stock control became more accurate and transaction records more precise. Time spent on manual coordination decreased.
It also freed up capacity for equivalent projects in manufacturing processses, as management were able to spend less time dealing with all these things and could focus on further improvement.