Can AI Create Real Value in Logistics Today?

From left: Markus Messmer( DHBW Ravensburg), Heiko Marquardt (EFESO Management Consultants), Dr. Norman Dziengel (Inpixon)

Event Recap: “Everything Better with AI in Transport Logistics?” – DHBW, BME & BVL

Few industries feel process pressure like turbine manufacturing. When takt (the maximum allowable time to produce one unit to meet customer demand) slips, entire schedules do.

At the RITZ Innovation Center in Friedrichshafen, Inpixon and EFESO Management Consultants took the stage to show how real-time data and AI can help logistics and manufacturing teams keep production flow stable, even in complex, high-mix environments.

The joint session by Dr. Norman Dziengel (Inpixon) and Heiko Marquardt (EFESO Management Consultants) was part of the event “Everything Better with AI in Transport Logistics?” hosted by DHBW, BME, and BVL on November 6, 2025. It focused on one central question: How can AI create real value in logistics today?

Participants of the “Everything Better with AI in Transport Logistics?” event, RITZ Innovation Center Friedrichshafen — with Dr. Norman Dziengel representing Inpixon.

The Strategic Context: Where AI Is Making an Impact in Logistics

EFESO opened the joint session with a view across the current AI application areas in logistics, from the supply network to in-plant operations:

• Supply chain transparency & ETA prediction: AI-driven forecasts and machine learning-based tracking
• Demand and replenishment forecasting: Real-time stock optimization
• Inbound and quality assurance: Visual inspection and anomaly detection with deep learning
• Intralogistics and material supply: Dynamic material planning, adaptive OTD control, and AI-based prioritization

Their core message was clear: AI needs structure.

Only when lean processes, clean data, and defined ownership exist can AI truly act as a lever for performance.

In the second part of EFESO’s input, the focus shifted to the operational foundation: Lean + Digital = AI-ready. Without standardized workflows, even the smartest model won’t stabilize a process. With structure in place, however, AI can amplify flow instead of complexity.

From Plan to Reality: Inpixon RTLS Measures What’s Really Happening

Next in the presentation, Inpixon demonstrated how real-time location systems (RTLS) turn visibility into control in turbine manufacturing, where complexity, product variance, and takt times are high.

The challenge:
Paper-based tracking, delayed feedback, and unsynchronized WIP data between ERP/WMS and the shop floor lead to unstable cycles and reactive planning.

The solution:
A live operational layer that connects the digital plan with the physical reality, tracking every cart, asset, and WIP item in motion.

RTLS: Turning Assumptions into Facts

• Location tags track transport carts, assets, and WIP across zones and stations
• E-paper displays show live order status and trigger route updates
• Time stamps and process time series provide factual process analytics
• Automatic data feedback into ERP/WMS closes the loop

Result:

Cycle and transport times are measured, not guessed. Processes become visible and deviations transparent, forming the factual foundation for AI-based decision-making.

askPixi Adds the Why and What’s Next

Built on that foundation, Inpixon’s AI layer, askPixi, delivers the next level: context, prediction, and recommendation.

"RTLS gives us the facts. AI connects them into a story. Suddenly, process data starts explaining itself — deviations are no longer just numbers, but insights that guide action. That’s how we move from measuring performance to actually managing it."
— Dr. Norman Dziengel, Senior Product Manager & Consultant, Inpixon

What askPixi learns:

• Dwell-time patterns by route, station, and shift
• ETA and OTD risks, as well as potential material shortages
• Root causes such as congestion, sequencing errors, or batch mismatches

What askPixi recommends:

• Material rerouting, reprioritization, replenishment, or shift adjustment
• Natural-language explanations such as: “Dwell time in Zone B exceeds target, delay likely on Line 2. Activate Route N1.”
• Writes back parameters like cycle buffers, route frequencies, or safety stock levels directly into ERP/MES systems

The impact:

• Stable takt and throughput
• Fewer unplanned shortages and manual interventions
• Early warnings instead of reaction
• Optional agent-based automation once processes stabilize

In short: RTLS makes the floor measurable. askPixi makes it understandable and actionable.

Three Practical Examples from the Session

1. Clearing a Bottleneck

Situation: Station B is blocked. Cart W-342 has been standing still for 38 minutes — the next takt is about to slip.

Data used:  Live cart positions and timestamps, zone identifiers, shift plan, takt targets, historical dwell and travel times, and current queue data.

What happens:  The AI automatically detects the bottleneck, compares live positions, travel times, and the shift plan, and instantly proposes three corrective actions:

1. Reroute other milkruns via side route N1
2. Temporarily move one operator from Station C to Station B for 45 minutes
3. Send one person to inspect cart W-342 and identify the root cause

What operators experience: “I can see directly on the display that my cart has a new route. No waiting, no questions, the line keeps running.”

Result:  The handover stays on schedule, stress decreases, and the response happens in real time.

2. Bundling Deliveries Instead of Single Runs

Situation:  Five bins are running empty one after another — previously this meant five separate delivery runs.

Data used:  RTLS-based Kanban consumption messages, warehouse stock data, live milkrun positions, travel times, and aisle utilization data.

What happens:  The AI detects material needs through RTLS signals and automatically combines them into one optimized delivery tour.

What operators experience:  “My e-paper display instantly shows the new pick list and route. I only drive once — and nothing stands still.”

Result:  30–40% less driving time, no material shortages, and a stable line supply.

3. Prioritizing When Multiple Orders Are Delayed

Situation:  Three orders are delayed. Resources are insufficient for all, and the supply chain is at risk of stopping. A chip supplier has halted deliveries due to a rare-earth embargo.

Data used:  Revenue, customer type (new or existing), estimated delay, material delivery status, global political situation, and supplier portfolio.

What happens:  The AI automatically calculates an impact score and proposes an optimized priority sequence.

• A chip shortage threatens production from week 15 onward.
• The system identifies an alternative supplier and immediately initiates a replacement order.

What operators experience:  “I can see on the e-ink display that Order C has top priority — so that’s where I go first. No guessing, no discussions.” Also, “The replacement delivery is triggered immediately to prevent the upcoming chip shortage.”

Result:  Decisions are transparent, customer priorities are clear, workflows stay calm, and supply chain continuity is maintained.

Expected Effects Across Implementations

• OTD stability: up to 30% increase
• Takt stability: up to 70% increase
• Expedites: up to 50% decrease
• Manual interventions: up to 70% decrease

While exact values vary by setup and process discipline, the pattern remains consistent: Once the event-to-action loop is closed, operations become calmer, faster, and more predictable.

Takeaway

EFESO set the stage by showing how lean and digital foundations prepare organizations for AI.
Inpixon showed what happens when those foundations meet real-time intelligence:

RTLS measures the facts. askPixi understands them. Together, they act.

The result is a shift from firefighting to flow control with data-driven stability, visible causes, and actionable foresight.

As one participant summed it up:

“When visibility meets causality, the shop floor finally runs calm and the plan becomes believable.”

Acknowledgments

Thanks to DHBW, BME & BVL, and all speakers and participants for the open, practice-driven dialogue.
Special thanks to EFESO Management Consultants, especially Heiko Marquardt, for connecting strategic perspective with operational depth.

Contact:

• Dr. Norman Dziengel — Senior Product Manager & Consulting, Inpixon
• Heiko Marquardt — Expert Director, EFESO Management Consultants