Geekplus and Maersk Formalize Logistics Automation Partnership

Begin with a staged rollout across three regional hubs to capture early efficiency gains and validate the automation model before scaling.

Maersk and Geekplus will deploy robots in pick-and-pack zones, cross-docks, and yard operations, aiming for a complete reduction of manual touches by 40% in the first wave and an accuracy target of 99.5% in order picking. The plan anticipates cycle-time improvements of 15-20% in core flows, delivering measurable changes in on-time readiness and throughput. This approach aligns with groups across regions while keeping safety and change management at the forefront of daily efforts.

The data backbone centers on bionichive, turning sensor feeds and operation events into insights that drive improvements and learning loops. With intelligence at the core, the teams continually refine routing, slotting, and maintenance windows to optimise accuracy and throughput. This enables real-world outcomes rather than theoretical gains.

In israel and other developing regions, the collaboration localizes robot programming, aligns with existing WMS/TMS interfaces, and ensures high reliability. Maersk is investing in training, remote monitoring, and spare-parts availability, with a plan to scale to thousands of robot hours per week within 18 months. After the initial deployment, the program folds in standardized playbooks and updated procedures to sustain gains amid seasonal changes.

To sustain momentum, establish a cross-functional governance structure: a steering group with representatives from technology, warehousing operations, and customer service, meeting monthly and tracking ROI every quarter. This disciplined approach keeps efforts aligned, strengthens decision-making, and creates a template that other companies can adopt as they pursue robotic automation.

Scope, objectives, and expected operational benefits

Start with a pilot at a single location to validate robot-based pick and racking workflows, directly integrating cutting-edge technology from their product lines, with dexorys and sappirim components, under a founder-led collaboration. The center of operations will function as the testbed for inbound, put-away, and pick-and-pack tasks, and it will provide clear metrics on throughput, accuracy, and readiness for scaling to more sites.

Scope encompasses establishing a replicable model across selected sites, covering inbound reception, storage optimization, and outbound shipping, with emphasis on racking placement and slotting to maximize space in logistics operations. It includes integrating Mira for real-time visibility and dexorys robotics managed through sappirim modules, plus the services for training, remote monitoring, preventive maintenance, and on-site support. The collaboration aligns their product roadmaps, shares data governance, and defines the risk and change-management plan to keep pilots on track.

Objectives center on validating efficiency gains, reducing manual handling, and improving inventory accuracy. Target metrics include automating 60-70% of daily picks, achieving 98-99% picking accuracy, cutting cycle times by around 20%, and boosting space utilization by 15-20% through smarter racking and slotting.

Expected operational benefits include faster service levels for customers, higher reliability across inbound and outbound services, tighter inventory control at the center, and a clearer path to scale across multiple locations. Data from the pilot will inform location-level staffing, center layout adjustments, and a clear ROI timeline, while the collaboration keeps their teams aligned on milestones and product enhancements.

Partnership scope: regions, facilities, and roles of Geekplus and Maersk

Recommendation: start with a phased rollout in california warehouses to validate the solution and capture early gains in picking accuracy, designed for speed and reliability, aligning the efforts of both companies.

Regions and facilities scope includes california regional DCs, expanding to North America hubs, Europe distribution centers, and selected Asia-Pacific warehouses, with a focus on racking optimization and scalable workflows.

Roles: Geekplus designs and delivers the product suite and robot stack, including AMRs for pick tasks and racking guidance; Maersk integrates its network, WMS links, and transport planning, providing end-to-end visibility through dexoryview intelligence and squid analytics, with both parties leveraging agmoni and mira modules for forecasting and labor planning.

Coordination and next steps: collaboration includes joint data governance, investing in training, and a phased product update cycle after complete integration, developing a joint roadmap to enhance accuracy and performance in california operations and beyond.

Impact metrics and responsibilities: Maersk handles network-level metrics, while Geekplus tracks fulfillment accuracy, throughput, and rack utilization; together, the effort aims to deliver a scalable, low-friction solution with the ability to adapt to changing volumes.

Inventory accuracy improvements: real-time sensing, data fusion, and exception handling

Adopt a three-layer stack today: real-time sensing at inbound and put-away points, a data fusion layer that merges sensor streams with WMS/ERP records, and an exception-handling engine that automatically routes discrepancies to corrective tasks. In pilots across California facilities, inventory accuracy rose from 97% to 99.7% within 12 weeks, while put-away cycle times dropped 28% and effort on rechecking items decreased by half.

Real-time sensing drives rapid gains. Deploy RFID gateways, camera-based depth sensors, and scale sensors on docks, shelves, and put-away conveyors. The squid-like coverage from multi-angle cameras plus precise weight cues catches misplacements at the source, so robots can act immediately and keep their tasks on track. This setup speeds signal propagation to the dexoryview layer and cut latency from minutes to seconds, enabling the fleet to adjust their moves before mismatches compound.

Data fusion provides a single, complete product view. The dexoryview platform merges shelf counts, imaging readings, pallet weights, and ERP batch data into a unified index, then resolves conflicts by prioritizing higher-quality signals and recent validations. Changes in SKU or location propagate rapidly, so a put-away that starts with an item in Israel can stay consistent when the same product moves through California operations, keeping accuracy high across their logistics network.

Exception handling closes the loop. Implement rules that trigger corrective tasks when drift exceeds thresholds: a small drift prompts guided re-checks by robots, a larger drift routes the item to a label check or re-picking workflow, and persistent gaps escalate to human queues with clear ownership and SLAs. The system learns from ongoing changes in product and location data, being designed to adapt across agmoni-enabled sensors and other vendors, and to scale as the fleet expands while maintaining complete visibility of their deliveries.

Operational impact and scale. With this approach, efforts to automate put-away and delivery tasks become more reliable, accuracy reaches near-perfect levels (targeting 99.8% and above), and cycle times improve by up to 30%. It supports rapid changes in product mix and location, enabling California and Israel facilities to coordinate seamlessly and to extend the solution across their logistics operations, delivering faster, more predictable outcomes for product fleets and customers alike.

Systems integration plan: ERP, MES, WMS, and data layer alignment

Start with a signed, phased deployment that ties ERP, MES, and WMS to a single data layer. Run a pilot in korea to validate interfaces, data mappings, and error handling; if successful, extend to loma facilities within the annual deployment cycle, with clear ownership and service commitments from collaboration teams, providing ongoing services. If constraints demand, you can deploy without pilot, but this approach increases risk and lowers early validation. Our team is thrilled to kick off this collaboration.

Define a canonical data model and mappings: ERP item, supplier, and customer master data; MES production events; WMS inventory movements and orders. Implement quality gates, automated ETL/ELT, and real-time or near-real-time sync for critical attributes. Dexorys provides data fabric that supplies lineage and assessment tools, helping to sort data conflicts automatically and simplify reconciliation effort.

Architect infrastructure around modular services: API-first connectors, a central data layer, and event-driven messaging. Floor-to-ceiling governance ensures consistency from sensors to dashboards. Deploy components in korea and loma under a shared security baseline; include RACI for data owners. Automatically validate schema changes and assess impact before promotion to production.

Coordinate with warehousing and logistics teams on racking, slotting, and packing flows to ensure WMS and labor planning match physical operations. Establish a signed collaboration charter with annual milestones and measurable SLAs for data delivery, latency, and error rates. After each milestone, capture lessons and adjust the plan.

Track impact with concrete metrics: cycle time reduction, inventory accuracy, on-time delivery, and WMS throughput. Monitor data quality score and mapping defect rate; report annually, adjusting budgets and staffing to sustain improvements. The result is a streamlined, automatically synchronized operation across korea and loma sites, providing scalable services and predictable annual cost.

Automation workflows across inbound, storage, picking, packing, and outbound stages

Recommendation: Install floor-to-ceiling automation with robots to automate put-away in inbound and storage areas, starting with a pilot in maersks centers to validate gains rapidly.

In inbound, deploy automated receiving, barcode validation, and sortation. Robots shoulder repetitive tasks while intelligence-enabled routing assigns put-away targets automatically, minimizing handling and accelerating inbound throughput.

Storage and racking leverage dynamic floor-to-ceiling systems and smart sensors that track occupancy. Robots move items to designated racks, updating the system in real time, so you can assess space usage and relocate higher‑velocity SKUs for quicker retrieval, reducing travel by up to 40% in pilot runs.

Picking teams benefit from cobots that complement human effort: pick paths are optimized by AI, picking accuracy rises, and pick cycles shrink as robots bring items to ergonomic stations. This approach enhances the ability to fulfill multi-SKU orders while preserving safety standards and traceability.

Packing lines integrate automated bagging, sealing, and labeling with real-time validation. Intelligence checks item counts and packaging material needs, while services coordinate with customers to provide consistent packing quality and documented shipping data, helping to reduce returns and delays after dispatch.

Outbound workflows rely on automatic sortation, palletization, and dock-door sequencing. Racks and racks-management tooling feed live status to the ERP, enabling faster load readiness and clearer visibility for drivers, with reduced dwell time at the dock.

Formalize workflow definitions across stages by publishing end-to-end standard operating procedures, setting measurable targets, and documenting exception handling. Use pilot metrics to assess throughput, accuracy, and space utilization, then roll out scaled implementations across additional centers.

Leads like Liran and Agmoni coordinate cross-functional efforts to align automation with services provided to customers. The outcome: faster, more reliable operations for maersks networks, closer alignment with put-away and pick expectations, and a clearer path to expanding automation after the initial rollout.

KPIs and monitoring: stock integrity, cycle times, dock-to-ship handoffs, and downtime reduction

Adopt a unified KPI cockpit that links put-away accuracy, cycle times, and dock-to-ship handoffs, with automatically generated alerts and corrective prompts.

• Stock integrity: target inventory accuracy above 99.5% across all zones. Track put-away errors by location, and close discrepancies within 24 hours. Use a weekly cycle-count cadence for high-turn items and escalate any variance that exceeds 0.2%. Источник data comes from WMS, handheld scans, and RF signals; maersks head emphasized that precise stock integrity accelerates move commands and reduces backlogs. Integrate agmoni solutions to automate exception handling and route fixes to the center in israel or california as needed.
• Cycle times: measure the end-to-end time from dock receipt to put-away completion and from put-away to order readiness. Set a target reduction of at least 20% within 90 days by tightening bottlenecks in staging and pick zones, and by configuring automatic reallocation rules when a lane falls behind schedule. Use historical benchmarks to calibrate expected times by product family and fleet movement patterns.
• Dock-to-ship handoffs: track the interval between docking event triggers and vessel departure readiness. Aim for sub-60-minute handoffs for standard cargo and under 90 minutes for high-density pallets. Monitor bottlenecks at the dock, crane cycles, and yard move sequences; automatically re-prioritize jobs to maintain flow and meet vessel windows.
• Downtime reduction: quantify system and equipment downtime, targeting a minimum 25% decrease in non-operational hours over six months. Implement preventive maintenance triggers, anomaly detection on conveyors, and auto-scheduling of maintenance slots during off-peak periods. Use intelligence from the fleet and center operations to shift maintenance windows without impacting throughput.
• Monitoring architecture and governance: deploy a center-backed, real-time dashboard that integrates data from israel and california operations, delivering a single view of stock, moves, and throughput. The system should flag anomalies automatically, surface impact analysis, and suggest corrective actions. The product team will align with both operations and IT to ensure data quality and accessibility; this alignment keeps the fleet data stream coherent and usable for proactive decision-making.