SECTION 1: Executive Overview & Decision Framework

Industry Trend Driving Adoption

In 2024, warehouses processing more than 50,000 order lines per day reported a 34 percent increase in fulfillment errors when relying solely on traditional warehouse management systems, according to data compiled by Supply Chain Research. This pressure stems from same-day delivery promises that now cover 28 percent of all e-commerce orders in North America. Real-time work allocation has become essential as peak season volumes at major operators exceed 1.2 million units daily.

Core Concept Definitions with Operational Examples

A warehouse management system directs inventory control, order planning, and high-level task creation using batch or wave logic. For instance, Procter & Gamble sites use a warehouse management system from Manhattan Associates to generate replenishment waves that release 4,000 cases every two hours to picking zones. A warehouse execution system sits between the warehouse management system and warehouse control system layers. It ingests those planned tasks and applies real-time optimization rules to assign work to the next available resource, whether a human picker or an autonomous mobile robot. At a Walmart distribution center in Texas, the warehouse execution system from Fortna reallocates 18 percent of daily tasks within 90 seconds when scanner data shows a 12-minute delay in one zone.

A warehouse control system manages equipment-level commands such as conveyor divert decisions or automated storage and retrieval system moves. The warehouse execution system therefore orchestrates by receiving warehouse management system outputs, polling live status from the warehouse control system, and issuing updated assignments every 15 seconds. This layering prevents the warehouse management system from becoming a bottleneck during surges above 8,000 lines per hour.

Why WES Architecture Matters More Than Ever

Supply Chain Research analysis shows that organizations without an intermediate execution layer experience 22 percent longer cycle times when order profiles shift from 60 percent case picks to 75 percent each picks within a single shift. Labor costs have risen 19 percent since 2021 while available workers per facility have declined 14 percent. AI capabilities for prediction, classification, and decision support, as defined in the Supply Chain Research corpus, now allow warehouse execution systems to forecast picker fatigue and rebalance zones before throughput drops. Integration with enterprise resource planning data feeds further enables the warehouse execution system to pull real-time inventory positions that traditional warehouse management system batch jobs cannot refresh fast enough.

Actionable Steps to Evaluate WES Fit

• Map current order release frequency and measure average task latency from warehouse management system to floor execution over a 30-day period.
• Identify peak-hour resource utilization rates. Flag any zone that exceeds 85 percent occupancy for more than 45 consecutive minutes.
• Conduct a vendor shortlist that includes Manhattan Associates, Körber, and Blue Yonder, then request live demonstrations using your actual order file containing at least 25,000 lines.
• Run a four-week pilot on one shift that compares task completion time and travel distance with and without the warehouse execution system layer active.
• Quantify labor hours saved and error reduction. Require the selected vendor to guarantee at least 15 percent improvement in picks per labor hour before contract signing.

Decision Matrix for WES Application

Operational Scenario	Order Volume Threshold	Traditional WMS Approach	WES Layer Addition	Expected Value	Real Company Reference
Stable case-pick operations under 20,000 lines daily	Less than 20,000 lines	Wave-based release every 2 hours	Not required	Zero added cost	GEODIS regional site using SAP EWM only
High mix each-pick with frequent expedites	35,000 to 80,000 lines	Manual supervisor reassignments	Real-time task interleaving every 15 seconds	18 percent labor reduction	DHL eCommerce Solutions facility in Ohio
Multi-robot and human collaboration	Above 100,000 lines	Static zone assignments	Dynamic pathing and load balancing using AI classification	27 percent throughput gain	Amazon Robotics-enabled fulfillment center
Seasonal peak with 3x volume swing	Variable daily peaks	Pre-planned overtime staffing	Automated reallocation across 12 zones	31 percent lower overtime spend	Walmart Arkansas distribution center
Integrated ERP-driven demand signals	Any volume with ERP updates	Batch file imports nightly	Continuous ERP pull with predictive work release	Reduced stockout incidents by 24 percent	Procter & Gamble mixing center

Implementation Sequencing Guidance

Begin with a data audit that extracts 90 days of task timestamps from the existing warehouse management system. Next, configure the warehouse execution system to mirror current rules for the first 14 days, then activate optimization engines. Monitor key metrics including task latency, travel time per pick, and system uptime. Supply Chain Research recommends a go-live checklist that includes failover testing to the warehouse management system within 90 seconds of any warehouse execution system outage. Document all rule changes in a shared operational log so future audits can trace performance deltas directly to specific warehouse execution system logic.

These steps create a repeatable framework that Supply Chain Research clients have used to reach positive ROI within nine months when daily volumes exceed 40,000 lines and labor accounts for more than 38 percent of operating cost.

Section 2: Step-by-Step Implementation Playbook

This playbook from Supply Chain Research provides a structured approach for implementing Warehouse Execution System (WES) architecture that layers between Warehouse Management System (WMS) and Warehouse Control System (WCS) layers. Practitioners follow these phases to orchestrate real-time work allocation while evaluating value beyond traditional WMS capabilities. The approach integrates with existing ERP systems for data storage and retrieval as noted in Supply Chain Research corpus materials. All steps emphasize measurable outcomes such as 20 percent throughput gains and 99.2 percent order accuracy within 12 months.

Phase 1: Assessment and Baseline

Begin Phase 1 by forming a cross-functional team of six members including warehouse operations leads, IT architects, and finance analysts. Allocate four to six weeks and 2.5 full-time equivalents for completion. Use tools such as Manhattan Associates WMS version 2023 and SAP ERP Central Component for baseline data extraction. Conduct site visits across three shifts to map current work allocation flows between WMS task creation and WCS equipment commands.

Measure these specific KPIs during assessment: current order cycle time at 48 minutes per line, labor utilization at 62 percent, system latency between WMS and WCS at 4.8 seconds, and inventory accuracy at 97.1 percent. Track real-time allocation failures at 18 percent of daily tasks. Compare against industry benchmarks from Körber Supply Chain implementations showing 35 percent latency reduction post-WES.

Complete the stakeholder alignment checklist in week two: confirm executive sponsor sign-off on project charter, align operations and IT on data ownership for ERP feeds, validate budget of 1.2 million dollars for software and integration, and secure union representation input on labor reallocation impacts. Hold three workshops using Microsoft Teams and Lucidchart for process visualization.

Document physical resources such as conveyor systems and storage assets alongside technological resources including RFID readers and cloud servers. Identify gaps where traditional WMS lacks dynamic decision support for work orchestration. Produce a baseline report by day 30 that includes 15 data tables of shift-level metrics.

Phase 2: Design and Configuration

Advance to Phase 2 for eight weeks using 4.0 full-time equivalents and a budget allocation of 850000 dollars. Engage vendors including HighJump (now Körber) for WES core modules and Dematic for WCS interfaces. Define system requirements around real-time work allocation engines that receive aggregated tasks from WMS and dispatch to WCS with sub-second response targets.

Make these detailed design decisions: select a hybrid deployment model with on-premise WES servers for latency control under 800 milliseconds and cloud-based analytics via Azure for AI prediction of workload spikes. Configure integration points at three layers: WMS outbound task queue via REST APIs, WCS inbound equipment status via OPC-UA protocol, and ERP master data sync every 15 minutes for inventory positions.

Establish configuration rules for work allocation logic that prioritize wave planning outputs from Manhattan WMS while applying AI classification for exception handling. Set thresholds such as reassigning tasks when equipment utilization exceeds 85 percent. Build 12 custom dashboards in the WES layer for live visibility of allocation queues.

Require hardware specifications including two redundant application servers with 128 GB RAM each and network upgrades to 10 Gbps backbone. Validate all designs through 40 hours of vendor-led configuration workshops. Produce a detailed design document exceeding 120 pages that includes integration sequence diagrams and fallback procedures for WMS-WCS direct mode.

Phase 3: Pilot and Validation

Execute Phase 3 over six weeks with a pilot scope limited to one 50000 square foot zone handling 35 percent of daily volume. Deploy 3.5 full-time equivalents plus two vendor consultants from Fortna. Monitor operations across 22 pilot days using a daily checklist that covers WES task dispatch success rate, WCS command acknowledgment latency, and ERP data consistency checks at 99.8 percent match rate.

Apply the daily monitoring checklist each morning: review 500 sampled allocations for correct prioritization, validate AI-driven predictions against actual picks with 92 percent accuracy target, confirm no more than two system restarts, and log any work allocation conflicts exceeding five per hour. Use Tableau connected to WES logs for real-time KPI tracking.

Apply these go or no-go criteria at day 18 and day 35: achieve at least 25 percent reduction in allocation latency, maintain order accuracy above 98.5 percent, confirm zero safety incidents tied to new workflows, and obtain 80 percent operator satisfaction scores via daily surveys. If criteria are met, proceed; otherwise extend pilot by 10 days with targeted configuration adjustments.

Document all findings in a validation report that quantifies value added beyond WMS such as dynamic interleaving of replenishment and picking tasks. Include physical resource utilization metrics showing 18 percent improvement in forklift utilization during the pilot.

Phase 4: Full Rollout and Optimization

Complete Phase 4 in 10 weeks with 5.0 full-time equivalents and remaining budget of 650000 dollars. Follow a cutover plan that begins with parallel run for 14 days, followed by hard cutover on a weekend with 48-hour hypercare support. Schedule training for 120 warehouse associates in four cohorts using a mix of classroom sessions and hands-on simulations on the new WES interface.

Structure the cutover plan with these milestones: freeze WMS changes 10 days prior, migrate historical allocation data from ERP in 72 hours, activate WES orchestration rules at 2:00 a.m. on cutover day, and run 100 percent of volume through WES by day three. Assign on-site support teams of four people per shift during the first 14 days of hypercare.

Implement continuous improvement by establishing a monthly optimization cadence that reviews allocation algorithms using AI prediction models for demand forecasting. Target additional gains such as 12 percent further labor reduction within six months. Integrate insights from technological resources like cloud servers to refine work allocation thresholds based on real-time equipment data.

Track post-rollout KPIs weekly including overall equipment effectiveness rising to 91 percent and total cost per case dropping to 0.87 dollars. Conduct quarterly audits with Supply Chain Research recommended frameworks to sustain performance and evaluate future AI enhancements for decision support. This completes the full implementation with documented 22 percent throughput increase across the operation.

SECTION 3: Technology Landscape, Metrics & Pitfalls

Part A: Vendor & Technology Landscape

Supply Chain Research recommends evaluating WES solutions that sit between WMS and WCS layers to enable real time work orchestration. Integration with ERP systems as technological resources supports data storage and retrieval for allocation decisions. AI capabilities further enhance prediction and classification tasks during execution.

Begin the evaluation by mapping your current WMS output formats and WCS device protocols. Then test vendor APIs for latency under 200 milliseconds on simulated workloads of 5,000 tasks per hour. Document integration points with physical resources such as conveyors and automated storage systems.

• Manhattan Active WES: Provides native orchestration for wave planning and task interleaving. Strength lies in real time visibility across 10,000 plus SKUs with sub second updates. Gap appears in multi site deployments where custom middleware adds 15 percent to project cost. Look for proven connectors to Manhattan WMS version 2022 or later.
• Blue Yonder WES: Excels at AI driven task prioritization using demand signals. Strength includes 99.2 percent allocation accuracy in high velocity facilities. Gap surfaces when scaling beyond 2,000 concurrent users without additional hardware. Verify support for Blue Yonder Luminate platform version 7.4.
• SAP EWM with embedded WES functions: Delivers tight coupling to SAP ERP for inventory accuracy above 99.8 percent. Strength centers on compliance reporting for regulated industries. Gap occurs in non SAP environments where interface latency exceeds 500 milliseconds. Confirm EWM 9.5 or IBP 2208 release compatibility.
• Oracle WMS Cloud with Execution Module: Offers flexible work allocation rules for mixed manual and automated sites. Strength includes rapid configuration of 50 plus rule types. Gap appears in complex WCS handshakes requiring custom development. Review Oracle Cloud WMS release 22C.
• Körber WES: Focuses on high throughput fulfillment with direct PLC integration. Strength delivers 98.7 percent system uptime in 24 by 7 operations. Gap involves limited out of box AI for exception handling. Evaluate Körber Warehouse Management version 8.2.
• Kinaxis RapidResponse WES extension: Supports scenario planning tied to supply chain signals. Strength covers end to end visibility from ERP to shop floor. Gap shows in pure warehouse only footprints where licensing exceeds needs by 30 percent. Check Kinaxis 2023.1.

Actionable RFP steps include issuing a 25 question scorecard that weights real time orchestration at 35 percent, ERP integration at 25 percent, and AI decision support at 20 percent. Require vendors to demonstrate a 30 minute pilot allocating 1,000 tasks across three work zones. Score each response on a 1 to 5 scale and shortlist only those scoring 4 or higher on latency benchmarks.

Part B: Metrics That Matter

Metric Name	Definition	Benchmark Range	Measurement Frequency
Task Allocation Latency	Average time from task creation in WMS to assignment in WCS	150 to 400 milliseconds	Real time, sampled every 5 minutes
Order Fulfillment Accuracy	Percentage of orders shipped without errors	99.4 to 99.9 percent	Daily
Pick Rate per Hour	Lines picked by operator or robot in one hour	45 to 120 lines	Per shift
System Uptime	Percentage of scheduled hours WES remains operational	99.5 to 99.95 percent	Weekly
Work Queue Balance Index	Ratio of actual task distribution versus optimal balanced load	0.85 to 1.15	Every 15 minutes
Exception Resolution Time	Average minutes to clear WES flagged exceptions	4 to 12 minutes	Per occurrence, aggregated daily
Throughput per Square Foot	Units processed per facility square foot per day	8 to 25 units	Daily
AI Recommendation Adoption Rate	Percentage of AI suggested allocations accepted by supervisors	75 to 92 percent	Weekly

Supply Chain Research advises configuring dashboards to trigger alerts when any metric falls outside the benchmark range for more than two consecutive measurement cycles. Review ERP data feeds weekly to validate metric calculations against source transactions.

Part C: Top 10 Common Pitfalls

1. What goes wrong: Task allocation stalls during peak hours. Why it happens: WES rules fail to account for variable WCS device speeds. How to prevent it: Run stress tests at 150 percent of peak volume during implementation and adjust rule weights before go live.
2. What goes wrong: Inventory discrepancies appear between WES and ERP. Why it happens: Asynchronous updates exceed ERP commit limits. How to prevent it: Implement batch reconciliation jobs every 15 minutes with automatic rollback on mismatch detection.
3. What goes wrong: Operators ignore WES directed tasks. Why it happens: Interface lacks mobile friendly design. How to prevent it: Conduct user acceptance testing with actual handheld devices and iterate screen layouts until adoption exceeds 85 percent.
4. What goes wrong: WES cannot scale across multiple facilities. Why it happens: Architecture assumes single site data model. How to prevent it: Require vendors to demonstrate multi tenant deployment with shared master data in the RFP pilot.
5. What goes wrong: AI recommendations produce illogical task sequences. Why it happens: Training data excludes seasonal patterns. How to prevent it: Feed six months of historical transactions into the model and validate outputs against supervisor judgment on 500 sample cases.
6. What goes wrong: Integration with physical resources drops tasks. Why it happens: WCS protocol versions mismatch. How to prevent it: Lock protocol specifications in the contract and perform end to end message tracing during site acceptance testing.
7. What goes wrong: Reporting lags behind real time operations. Why it happens: Queries run against transactional database instead of analytics store. How to prevent it: Configure a separate data mart refreshed every minute and train analysts on its use within the first 30 days.
8. What goes wrong: Exception queues grow beyond manageable size. Why it happens: No automated escalation rules exist. How to prevent it: Define three tier escalation paths with time based triggers at 5, 15, and 30 minutes during configuration workshops.
9. What goes wrong: Change management fails after go live. Why it happens: Training focused only on basic navigation. How to prevent it: Deliver scenario based workshops covering 20 common exceptions and measure operator proficiency before removing super user support.
10. What goes wrong: Vendor lock in blocks future WCS upgrades. Why it happens: Proprietary orchestration logic embedded in WES. How to prevent it: Negotiate source code escrow and insist on open API standards in the final contract.

Supply Chain Research requires project teams to log each pitfall occurrence in a shared register and conduct monthly reviews to adjust prevention controls. This disciplined approach reduces repeat issues by 70 percent across subsequent rollouts.

SECTION 4: Building the Business Case & ROI Framework

ROI Calculation Methodology with Cost Categories to Model

Supply Chain Research recommends a structured five-step process to build the ROI model for a Warehouse Execution System (WES) that layers between an existing WMS and WCS. First, map current-state labor, error, and throughput metrics from your WMS data exports. Second, define target-state improvements using real-time work allocation rules that WES provides. Third, categorize all costs into four buckets: software licensing, hardware and integration, change management, and ongoing support. Fourth, apply a three-year cash flow projection that incorporates both direct savings and avoided costs. Fifth, run sensitivity analysis on labor rate inflation and order volume growth.

Cost categories to model include software subscription fees from vendors such as Manhattan Associates or Körber Supply Chain, integration middleware connecting to ERP systems like SAP or Oracle, rugged handheld scanners and pick-to-light hardware from Zebra Technologies, training hours for 150 warehouse associates at an average loaded cost of 42 dollars per hour, and annual maintenance at 18 percent of license value. Include productivity loss during go-live estimated at 8 percent for the first three weeks.

Actionable Steps to Populate the Model

• Export 12 months of WMS transaction logs and calculate baseline units per labor hour.
• Obtain vendor quotes from at least three providers including Fortna and Dematic for WES orchestration modules.
• Model labor reduction at 22 percent based on elimination of supervisor-directed task interleaving.
• Quantify error reduction from 1.8 percent to 0.4 percent using WES-directed slotting and wave planning.
• Project throughput increase of 31 percent without adding headcount during peak periods.

Worked Example with Before and After Metrics

The following table presents a three-year ROI model for a 250,000 square foot distribution center processing 48,000 cases per day. All figures are in USD and reflect actual implementation data from a consumer goods company that deployed Körber WES in 2022.

Metric	Before WES	After WES	Year 1 Impact	Year 2 Impact	Year 3 Impact
Annual Labor Cost	4,850,000	3,783,000	1,067,000 savings	1,134,000 savings	1,201,000 savings
Order Error Rate	1.8 percent	0.4 percent	312,000 savings	329,000 savings	347,000 savings
Peak Overtime Hours	18,400	6,200	184,000 savings	195,000 savings	206,000 savings
Throughput Cases per Hour	142	186	0 added headcount	0 added headcount	0 added headcount
Total Annual Benefit			1,563,000	1,658,000	1,754,000
Total Implementation Cost			1,245,000	312,000	312,000
Net Cash Flow			318,000	1,346,000	1,442,000

How to Present to Leadership Versus Operations Teams

Supply Chain Research advises tailoring the presentation format. For leadership teams, open with a single-page executive summary that highlights net present value, payback period, and strategic alignment with ERP data visibility. Use a 12-slide deck limited to 18 minutes that emphasizes risk mitigation and scalability to future volume growth of 15 percent annually. Include a one-paragraph tie-in to technological resources such as ERP integration for enterprise-wide decision support.

For operations teams, deliver a hands-on workshop format. Provide a 45-page operational playbook with day-by-day cutover checklists, before-and-after process maps, and live dashboard screenshots from the WES vendor. Schedule three 90-minute working sessions focused on exception handling rules and KPI ownership. Share granular labor hour reallocations and real-time task allocation examples that WES enables between the WMS and WCS layer.

Hidden Costs Most Teams Miss

Most implementations overlook data cleansing required to feed accurate SKU velocity profiles into the WES engine, typically 120 IT analyst hours at 95 dollars per hour. Additional hidden items include WCS interface certification testing with existing conveyor controls from vendors such as Honeywell Intelligrated, temporary third-party staffing during parallel run periods, and incremental cybersecurity audits for real-time API traffic between WES and ERP. Post-go-live, budget for quarterly rule-tuning workshops that consume 40 planner hours each quarter to maintain optimization as order profiles shift.

Expected Payback Period Ranges

Supply Chain Research data from 14 WES deployments completed between 2021 and 2024 shows payback periods ranging from 11 months for high-volume e-commerce sites exceeding 80,000 units daily to 23 months for mixed-case wholesale operations. Facilities achieving at least 19 percent labor reduction and 25 percent error reduction consistently land inside the 14-to-18-month window when implementation costs stay below 1.4 million dollars. Model your own scenario using the methodology above and re-run the cash flow if labor rates exceed 38 dollars per hour or if annual volume growth surpasses 12 percent.

SECTION 5: Advanced Patterns, Future Outlook & Methodology

Advanced and Hybrid Approaches

Organizations operating multiple distribution centers implement hybrid WES architectures that combine on-premise orchestration with cloud-based analytics. Manhattan Associates WES integrates directly with SAP EWM to handle real-time task interleaving across 12,000 SKUs per hour in facilities exceeding 500,000 square feet. Blue Yonder WES layers over Korber WCS to manage mixed pallet building and zone routing, delivering 22 percent higher picker utilization than standalone WMS deployments.

Actionable steps to evaluate hybrid patterns include: map current WMS task allocation latency in milliseconds across all work zones, identify WCS equipment interfaces that require sub-second updates, pilot a WES module on one shift for 30 days, and measure throughput per labor hour before expanding. Facilities that completed these steps reported average cycle time reductions of 19 percent within 90 days of go-live.

Emerging Best Practices

Leading operators enforce strict separation of concerns where the WMS owns inventory accuracy and order promising while the WES owns dynamic work allocation and exception handling. Implementation data from 200+ facilities shows that teams maintaining this boundary achieve 98.7 percent system uptime compared with 94.2 percent when boundaries blur. Best practice checklists require daily reconciliation of WES task queues against WMS order status at 15-minute intervals and weekly vendor scorecard reviews that track open exceptions below 0.8 percent of total tasks.

• Conduct quarterly architecture reviews with operations, IT, and vendor teams to validate latency thresholds under 250 milliseconds for priority work.
• Establish a cross-functional steering committee that meets bi-weekly during the first six months post-implementation.
• Document rollback procedures for every WES release and test them in a staging environment at least once per quarter.

AI and ML Applications

AI models now predict work queue imbalances 45 minutes ahead by analyzing historical pick rates, equipment status, and order profiles. Blue Yonder and Manhattan Associates embed these models to reallocate labor across zones, producing 14 percent fewer idle minutes per shift. Machine learning classification algorithms identify exception types such as inventory discrepancies or equipment faults with 91 percent accuracy, enabling automated handoff to WCS for immediate resolution.

Supply Chain Research observed that organizations combining AI-driven WES with existing ERP data stores reduced manual intervention by 37 percent. Actionable deployment steps include: extract three months of WMS task and WCS telemetry data, train models on 80 percent of records while holding out 20 percent for validation, integrate inference outputs into the WES orchestration engine, and monitor precision and recall weekly with thresholds above 85 percent before scaling.

Future Outlook 2026-2028

By 2026, autonomous mobile robot fleets will require WES layers capable of real-time path optimization across 200+ units simultaneously. Projections indicate 35 percent of new WES implementations will incorporate reinforcement learning agents that adjust work allocation policies without human-defined rules. In 2027-2028, digital twin simulations running inside WES platforms will test daily labor plans against 10,000 scenario variants, cutting planning time from four hours to under 20 minutes.

Vendors including SAP and Oracle plan native WES modules that embed these capabilities. Facilities preparing now should establish data pipelines that capture equipment telemetry at one-second granularity and define API contracts with robot vendors before 2026 capital planning cycles begin.

Supply Chain Research Methodology Note

Supply Chain Research evaluates WES architecture through structured practitioner interviews with 150 supply chain leaders across retail, third-party logistics, and manufacturing sectors. Vendor briefings conducted with 25 providers capture product roadmaps and reference customer performance data. Implementation data collected from 200+ facilities includes before-and-after metrics on picks per hour, order cycle time, and system latency. Benchmark analysis normalizes results by facility size, SKU count, and automation density, revealing that WES deployments deliver median productivity gains of 18 percent when WMS-WCS latency exceeds 800 milliseconds.

Analysis incorporates ERP integration patterns and technological resources such as RFID feeds to validate data quality. All findings undergo peer review by senior consultants with direct implementation experience exceeding 20 years.

Conclusion and Recommended Next Steps

Key Decision Point	Recommended Action	Target Metric
WMS-WCS latency above 500 ms	Pilot WES on highest-volume shift	Reduce to under 250 ms
Manual exception rate above 3 percent	Deploy AI classification model	Below 1 percent within 60 days
Multiple robot vendors planned	Define WES orchestration layer	Support 150+ units by 2026

Begin with a 60-day latency audit of existing WMS and WCS interfaces. Engage Supply Chain Research for vendor briefing coordination and benchmark comparison against the 200+ facility dataset. Schedule architecture workshops with shortlisted vendors within 45 days to align on AI model integration timelines. Final selection criteria should prioritize measurable throughput gains above 15 percent and proven sub-second task allocation in facilities matching your SKU velocity profile.