GOLAS by Speech Technology Center and iVoice for voice-controlled picking

Speech Technology Center and iVoice developed an atomized voice system GOLAS for picking optimization in warehouse environments. Pickers who have to compile large complex orders will be able to receive voice instructions via a headset, which will replace paper- or tablet-based picking. The system currently operates in English and Russian languages.

Autonomous and frost-resistant

GOLAS enables hands free and eyes free operation for pickers, as they do not have to use a hand held terminal. It is based on a Voice Picking principle – the system dictates the list of all the items from the order to the worker via a headset. This is especially useful when working on a truck or in a warehouse with big-ticket items, which have to be carried with both hands. The employee is not required to press any buttons or use a touch screen, so even when working in the cold it is not necessary to take off the gloves.  

«Depending on the industry, the solution pays off in about 12-18 months. Warehouse pickers are able to work faster (efficiency increases up to 35%) and the chances of making a mistake decline: until the worker indicates that a required item is in the right place, the system is not providing him the next task (the amount of mistakes falls 2-3 times). Working with GOLAS is easy - it takes only 2-3 days to teach an employee how to operate with it, so even with the high employee turnover it won’t create a problem,» - comments Ilya Ivanov, the CEO of iVoice.

Mobile terminal of GOLAS is designed for intensive daily use. It is constructed with wear-resistant materials, adapted for use at low temperatures (on the street, in refrigerators, freezers), is equipped with powerful batteries and does not require constant wireless connection to the server.

The solution is suitable for optimization of operations in warehouses, retail shops, online retailers, food production, etc.

Voice-controlled picking process

GOLAS works as follows. The head of warehouse forms an order and uploads it into the management system (WMS or ERP). The order is then converted into a step-by-step instruction – i.e. how many items have to be collected from a certain section, rack and shelf.

The instruction is then sent onto a mobile terminal, fixed on the picker’s belt. A synthesized voice instructs the employee, for example, to take 30 units of product A from the second shelf of the rack A1. Picker then executes the task and notifies the system about its completion into the microphone – GOLAS then provides him with the following step. An employee can ask the system to repeat the task if she did not hear or forgot it. The level of noise does not affect the speech recognition capabilities of the system.

Speech Technology Center’s voice module synthesizes the speech in one of nine standard voices. However, a client can also order a unique voice developed individually.

«Speech Technology Center specializes in speech synthesis and recognition in complex environments. At the core of the voice control module of GOLAS lies a speaker-independent voice recognition technology. The system equally well recognizes speakers, regardless of their age, gender and even accent,» - says Alexey Zakrevskii, head of development at Speech Technology Center.


Speech Technology Center is a developer of innovative systems in multimodal biometrics, speech synthesis and recognition, multichannel recording, processing and analysis of audio and video data.

iVoice – a system integrator, developer and technological partner of Speech Technology Center, specializing in projects involving speech technologies, process optimization and logistic innovations.