Akustische Kommissionierung

What is acoustic order picking?

In acoustic order picking – often referred to as voice picking – the picker communicates with the warehouse management system (WMS) exclusively via speech. A headset delivers the target location, item number and quantity as synthetic or recorded voice instructions. The employee confirms each step by speaking code words or short digit sequences, which a speech recognition module within the system then processes. Paper, scanners and displays are eliminated entirely, making this a paperless method.

How does acoustic order picking work in practice?

The process typically works as follows: the WMS transfers a pick order to a voice middleware or a dedicated voice system. This converts the order data into voice output and transmits it wirelessly to the picker's headset. After each pick, the employee speaks a check code displayed on the shelf label – this confirms that they are at the correct storage location. The system acknowledges the pick and issues the next instruction.

Typical application areas include fresh food logistics, food retail and high-throughput distribution centers, where hands-free operation and speed are especially important. In noisy environments – such as cold stores or areas close to production – the speech recognition must be specifically trained to handle ambient noise. This is achieved through so-called speaker-independent or speaker-dependent models; the latter are calibrated to each individual employee's voice and in practice often achieve lower error rates.

What should you look for when choosing a solution?

When introducing an acoustic picking solution, you should carefully evaluate the following points:

• WMS integration: Voice systems communicate with the WMS via standardized interfaces or proprietary middleware. Close integration reduces latency and susceptibility to errors.
• Environmental conditions: Noise levels, temperature (deep-freeze storage places special demands on headsets and batteries) and the acoustics of the facility significantly affect recognition accuracy.
• Language model and multilingual capability: In international teams, the system must be able to handle different accents and languages.
• Training effort: Speaker-dependent systems require an initial voice training session per employee, which creates administrative overhead in high-turnover environments.
• Hardware ergonomics: Headset design, wearing comfort and battery life are critical to user acceptance during shift operations.

Acoustic order picking can be combined with other methods such as pick-by-light or mobile scanners (so-called multimodal approaches) to deploy the most appropriate medium depending on the situation. Whether the investment is worthwhile depends heavily on order volume, item structure and existing IT systems – a blanket recommendation is not possible here.