3.3 Motion-Mining®: Automated Data Collection with Sensors & AI
Motion-Mining® marks a paradigm shift in process analysis: While classic methods are based on manual observation or model-based assumptions, Motion-Mining® collects process data automatically in a real work environment. Modern sensor technology records movements and activities directly in the process, and artificial intelligence analyses the resulting amounts of data objectively and reproducibly.
Employees wear compact wearables for this purpose. Each on both wrists and an additional sensor in the belt clip. In addition, Bluetooth beacons are used, which enable anonymized movement data to be spatially assigned. In this way, positions, movement patterns, activities and interactions can be recorded over a defined period of time without influencing the workflow. Data is collected during regular work and thus depicts real process situations instead of theoretical ideal processes.
The particular added value of Motion-Mining® lies in the continuous transparency of complex processes: sequences, interruptions, waiting times and paths become visible and quantifiable, especially where conventional analyses reach their limits.
New: Transparency also for Vehicles
Since 2025, Motion-Mining® has gone beyond analyzing human activities. With sensor-based fleet analysis, vehicle movements and loads can now also be systematically recorded. In many companies, these vehicle movements have so far resembled a black box:
- What is the actual utilization rate of individual vehicles?
- Where do waiting times occur?
- How much do empty runs account for the overall movement?
- At what lift heights do the vehicles work along the process?
Without objective data, these questions usually remain unanswered.
Data is collected via a modular sensor set with up to three sensors per vehicle. The basic sensor, the so-called Vehicle Logger, records basic vehicle movements. In addition, two additional sensors can be attached to the fork to measure fork run times, lift heights and empty and full travel, for example. The vehicles are located indoors via Bluetooth beacons and outdoors via GPS.
The combination of movement, usage and position data creates a complete, data-based picture of the intralogistics fleet for the first time. Inefficiencies become visible, potentials can be quantified and well-founded decisions for process and fleet optimization are possible.
Significant visualizations are created from millions of data points: heat maps show focal points of movement and make routes visible, time analyses reveal hidden waiting times or bottlenecks. The result is an objective, data-based image of actual work reality. Motion-Mining® does not model a target state, but documents the actual actual state of the process.
The strengths of Motion-Mining® lie in the combination of scalability, objectivity and realism. Unlike traditional methods, many employees can generate anonymized movement data simultaneously over weeks or months without the need for analysts to be physically present. Data is collected under real working conditions with all their uncertainties, disruptions and variations — exactly where the true “bottlenecks” are hidden. Observer effects are completely eliminated, and the sheer volume of data enables statistically robust statements even about rare events or highly variable processes. Motion-Mining® opens up completely new opportunities for intralogistics. Waste can be quantified with unprecedented precision: How many kilometers do order pickers travel every day? Where do waiting times occur systematically? Which areas are overloaded, which are underutilized? Which vehicles are used for internal transport? What is the ratio of full to empty runs and at what lift heights do my vehicles primarily operate?
These questions can not only be answered, but can also be differentiated according to shifts, weekdays or job profiles. The speed of gaining knowledge significantly exceeds traditional methods, as initial assessments are often available after just a few days. The low intervention intensity is particularly convincing. Employees can continue their work undisturbed, production processes are not interrupted, and acceptance is typically high with transparent communication. Many employees even appreciate that it is finally becoming objectively visible where structural problems lie, instead of individual performance being questioned. Motion-Mining® thus creates a common factual basis for improvement discussions between management, works council and operational teams.
In addition, the method provides an excellent basis for continuous improvement and agile optimization cycles. The effectiveness of changes can be measured precisely, A/B tests of various layout variants are possible, and the development of KPIs over time can be seamlessly traced. In combination with other data sources, such as ERP, WMS or MES systems, holistic process analyses are created that go far beyond pure movement data and provide even deeper insights into the processes. The basis for even better analysis results.
Limitations and classifications of Motion-Mining®
An initial hardware setup is required for data collection: Wearables must be provided and worn for the analysis period. The system must be equipped with beacons for indoor localization. It is only through the active participation of employees that real movement data can be recorded during operation. But that is exactly enough.
The implementation effort is deliberately low. Thanks to a plug-and-play approach, no integration into existing IT systems is necessary. The collected data is either automatically transferred to the analysis dashboard via WLAN or, alternatively, exported via a docking station using a USB stick. Both variants are easy, quick to implement and practicable even in IT restrictive environments.
Consistent consideration of data protection and compliance is also crucial. Motion-Mining® only collects anonymized movement data and no personal performance or behavior profiles. The use of data follows clearly defined legal and organizational frameworks and is an integral part of every analysis.
One limitation in terms of content is that Motion-Mining® currently provides data and transparency, but does not yet generate automated suggested solutions. The analysis results require interpretation and must always be evaluated in a specific process context. A long journey can indicate an inefficient layout or deliberately outsourced quality tests. Waiting times can be an expression of waste, but they can also be necessary buffers to stabilize fluctuating processes.
This is exactly where it becomes apparent that process data alone does not replace process expertise — but it creates the necessary basis for well-founded decisions.
In order to help companies derive concrete improvement measures, MotionMiners is expanding its Motion-Mining® portfolio with the digital process consulting platform SOLUTIONS. It starts right where data-based transparency becomes a need for decision-making. Using an intelligent search, behind which a large language model works, users can describe their specific challenges or filter them specifically for typical process problems — such as “long order picking” or “high ergonomic load.”
The platform specifically leads to so-called solution pages, which present a wide range of possible optimization approaches. These range from organizational measures to layout and process adjustments to technical solutions such as driverless transport systems, new zoning logic, exoskeletons or IT-based optimization of the picking sequence.
The solutions presented are based on MotionMiners' wealth of project experience from industry, trade and logistics. Each solution page transparently describes which process situations a measure is suitable for, which requirements must be met and which improvement potential is realistic. In addition, several providers are connected so that users can compare, evaluate and contact solutions directly.
SOLUTIONS thus acts as a neutral marketplace for process improvement and the platform makes use of solutions that come from the growing partner ecosystem. The platform significantly reduces the hurdle between identifying problems and finding solutions: Instead of one-sided market research or immediate consulting projects, companies receive well-founded guidance on possible courses of action within a short period of time.
Automated solution suggestions in the narrower sense have therefore not yet been fully implemented. But SOLUTIONS marks the next development step for MotionMiners: the systematic combination of data-based analysis, proven process expertise and structured access to specific optimization measures.