Executive Summary

Mavarick partnered with a global automotive OEM through the Digital Catapult Automotive Cluster Accelerator to apply its supply chain intelligence platform to tackle challenges with the movement of their goods, specifically the returnable transit packaging that move between plants, suppliers and carriers to keep production lines running.

Even though the OEM's supply chain management was working, there was an invisible cost attached to it. Issues were going undetected for days, manual processing was consuming hours of planner time per week, and costs were quietly accumulating across emergency spend, stagnated inventory and supplier discrepancies. Mavarick connected its platform to the OEM's existing data to monitor packaging flows across supplier loops and surface problems before they reached the line.

Key Takeaways:

• 90%+ anomaly detection rate
• Planner awareness time reduced from up to 72 hours to under 1 hour
• 85%+ of anomalies had a root cause automatically assigned
• Weekly data processing time reduced from up to 17 hours to 1 hour
• €150,000+ in packaging cost savings identified in just the first two weeks of roll out

Global Automotive OEM

The Global Automotive OEM has a manufacturing network across multiple plants, with returnable packaging moving constantly between a regional supplier base and production facilities. Managing that packaging pool, across multiple suppliers, carriers and inventory positions simultaneously, created a visibility and control challenge that manual processes could not keep pace with.At this scale, small inefficiencies compound quickly. Containers that sit idle tie up working capital. Supplier shortfalls that go unnoticed trigger emergency procurement. Discrepancies between what suppliers report and what is physically in the loop create gaps that planners cannot act on because they cannot see them. The cost does not announce itself, it builds in the background while the line keeps running.

The Challenge

Even though the packaging loop was working, there was a structural cost attached to it,  one that wasn't visible until the data was being systematically monitored

Problems surfaced too late: Issues were typically not flagged until suppliers raised them. By that point, emergency purchases had already been triggered to save production schedules. The window to intervene at low cost had closed before anyone inside the OEM knew there was a problem.

Root causes went unresolved: When issues were found, the immediate priority was containment, getting the line protected. Understanding why the problem had occurred came second, and often not at all. Without that analysis, the same issues kept recurring across the same supplier loops, generating the same costs repeatedly.

Manual processing was absorbing planner capacity: 100’s supplier queries were being handled manually every week, consuming up to 17 hours of planner time. That is time spent on data chasing and supplier clarification rather than on decisions that actually require human judgement. The administrative burden was not occasional, it was structural.

Costs were invisible:  Emergency packaging purchases, expedited freight and inventory discrepancies were being recorded as isolated operational events. Without traceability back to their source, there was no basis for understanding the true cost of running the loop reactively, or for making the case for change.

Inventory positions could not be trusted: In some cases, what suppliers reported, what carriers recorded and what the system showed were diverging by more than 20%. Planners were making decisions based on data that did not reflect physical reality, and had no reliable way of knowing when that was the case.

How Mavarick Addressed the Challenge

Mavarick defined five capabilities the platform needed to deliver from the outset. These shaped how the solution was designed and how success was measured throughout the pilot.

1. Detect - Provide visibility across all suppliers and loops and identify anomalies early
Mavarick's platform connected to the OEM's existing systems, pulling in data across inventory, shipments and supplier behaviour. The platform established performance baselines across the packaging loops and monitored against them in real time. When something deviated from expected norms, it was flagged immediately.

2. Respond - Enable rapid planner awareness with suggested actions
‍When the platform detected an anomaly, it automatically assigned a root cause alongside the alert. Planners knew what had gone wrong and why — without any manual investigation. That meant they could act on the problem straight away, not spend time diagnosing it first.

3. Prevent - Identify shortages and leakage before impact and assign root causes
‍Rather than waiting for a shortfall to become a shortage, the platform identified packaging gaps and forecast risks upstream — before they reached the production line. Root cause assignment meant those risks could be addressed at source, not just contained in the moment.

4. Improve - Drive measurable reductions in emergency packaging, expedites and pool imbalances
‍With problems surfaced earlier and root causes identified, the OEM had what it needed to reduce the recurring costs that had been accumulating quietly. Emergency procurement, expedited freight and pool misallocation could be addressed systematically rather than managed reactively case by case.

5. Sustain- Reduce manual burden and embed scalable, repeatable control
‍The platform brought the OEM's weekly data processing time down from as much as 17 hours to 1 hour. Supplier queries that had previously required manual handling were managed through the system directly. The result was a control model that did not depend on planner capacity to function, and one that scales as the supplier network grows.

Results and Impact

The pilot ran across a selection of strategic and challenging supplier loops:

• 90%+ of known anomalies in the data were detected by the platform
• Planner awareness time fell from up to 72 hours to under 1 hour
• 85%+ of detected anomalies had a root cause automatically assigned
• Weekly data processing time fell from up to 17 hours to 1 hour
• €150,000+ in packaging costs savings identified

This is a structural control upgrade with financial upside, not a reporting or visibility enhancement. 

Lessons Learnt

• A working loop is not the same as an efficient one: Even though the packaging loop was functioning, it was carrying a hidden cost. That cost only became visible once the data was being systematically monitored.‍
• Early detection is where the financial value sits: Finding a problem after a supplier has raised it leaves very few options. Getting ahead of it, before a shortfall becomes a shortage, is what prevents the cost from being incurred in the first place.‍
• Knowing why matters as much as knowing what: Identifying a problem is only useful if it leads to fixing the cause. With the majority of anomalies assigned a root cause automatically, the OEM could address issues at the source rather than managing the same symptoms repeatedly.‍
• Manual workload is a risk, not just an inconvenience: up to 17 hours a week on supplier queries is capacity not available for decisions that actually need human judgement. Reducing that to 1 hour changes what the team can do.

‍