Frequent Asked Questions

New to AllRead? Search for the most frequent asked questions or contact our support team.

General questions

What is a valid reading?

A valid reading is a unique reading with a confidence value over a set percentage, which is set by the client. For example, if the client sets a confidence value of 0.9 (90%), all unique reads over this value will be considered valid.

Is it possible to count the same reading twice?

No. “Reading” is understood as the code of an asset. It is not counted twice if the same code is read by two synchronized cameras. 

Does the software work offline?

Yes. The software does not need internet connection to read and process the images. However, an internet connection is required to receive automatic updates.

Does the software receive the updates automatically?

Yes. Clients who use either the SaaS solution, or on-premise solution, will have access to the latest readers, patches and new features. Clients using the on-premise solution will need to restart the ARS Docker container to receive these updates. 

Does your solution provide the OCR recognition only or also the GUI and data processing, storage, tracking, history, statistics, etc.?

A dashboard is available on the SaaS solution to all clients. Clients have access to current and historic reading data that has been processed in SaaS. 

On-premise clients will have access to a stateless dashboard where they can see a live feed of the readings.

Does your system need a gate or roof for your system to work?

We do not provide hardware. We use the fixed cameras or mobile devices of our clients. Until today we never needed any gate or roof. Sometimes a post is installed for the camera to be set at the right strategic place and height.  

Do you need an external trigger or signal that a dolly train is approaching or does your recognition system work continuously?

Our software can process continuous streaming and register all spotted container codes. Nevertheless, this is only possible if the software is installed on a physical server or inside the cameras (edge computing), to avoid too high costs (and latency) of cloud data processing. Therefore, it would imply additional hardware costs. 

We can best work with triggers: the camera activates itself (integrated motion detection) or is activated by an external sensor. A short video is then captured and sent to be processed by the software (on-premise / cloud server / SaaS), and the response is sent to the information webhook URL as JSON (or another format if needed).

In short, how does the integration of AllRead ARS work with the client´s server?

To process streams, ARS needs access to the client’s camera’s RTSP or MJPEG feed from either the on-premise platform, or cloud platform (hosted in AWS). After obtaining a valid reading, ARS launches an event that can either send to the client’s own API (webhook URL) or store in a database. 

Alternatively, to process images and videos, the client needs to send ARS files, after processing these readings are returned immediately to the requester.

Does the company offer a maintenance or remote support for the software?

Technical support is included in the license for all reading features, logins, dashboard… This does not include any infrastructure (servers) or hardware (cameras) maintenance. 

In case of an on-premise installation, the license includes a yearly model retraining using the onsite processed data. This will help further increase accuracy.

Can’t find the answer to any question?