Machine Learning vision at the edge of the network using LoRaWAN radios
How do you get started with an AI camera? The SenseCAP A1101 is a revolutionary IoT device that combines advanced Edge AI image analysis with long-distance LoRaWAN connectivity. It can be trained to perform various tasks, such as reading analog meters and transmitting data in an encrypted format. With LoRaWAN 's open radio bandwidth, users can use the device without operator fees.
What distinguishes the Seeed AI SenseCAP A1101 vision sensor?
SenseCAP A1101 is a state-of-the-art IoT camera for analyzing at the edge of the network (Edge AI) an image, and then relaying the collected data further through long-range LoRaWAN technology. The A1101 can be successfully trained to handle multiple activities requiring different scenarios. With it, we will be able to read values from all analog meters, convert them into digital format and transmit them in coded payload via LoRaWAN radio. Importantly, LoRaWAN operates on the open radio band, thus we will not have to pay any fees to the operator in order to use our device.
Training AI models on IoT camera
The device comes with software installed right from the manufacturer - two AI models are available for recognizing people's silhouettes and for reading data from analog meters. However, we can implement any AI model to achieve the expected functionality. By using TinyML models, developers of edgewise AI applications will be able to create applications for any purpose. Seeed SenseCAP A1101 is an excellent and low-cost platform for building such applications. On the other hand, thanks to the use of LoRaWAN technology, we will be able to implement projects with up to thousands of such devices spread over a very wide area.
Adaptable for use in all conditions
One of the biggest pluses of the described solution is that it can be implemented in virtually any conditions. Thanks to the IP66 waterproofness of the housing, the device will be possible to implement in outdoor conditions. The device is also capable of operating in extreme conditions: from -40°C to +85°C. In a typical application of this device, the battery will be able to last approximately as long as 9.5 months, which will significantly reduce the maintenance cost of the device.
How to get started with AI camera?
SenseCAP A1101 is a product that requires additional pre-adaptation for implementation:
- As a first step, you can test the SenseCAP A1101 with the two demonstration models provided in the product (human detection and meter recognition) and become familiar with the Vision AI application in a simple way;
- Collecting images using SenseCAP A1101 and generating the dataset through a development platform (e.g., RoboFlow);
- Train and generate custom models using external AI training tools and implement them in SenseCAP A1101;
- Work with an AI partner to generate datasets, train models, develop algorithms, etc., in order to obtain personalized and deployable models for specific applications and implement them in SenseCAP A1101.
Typical examples of using SenseCAP A1101
To list some of the ways in which the described device can be used, you can point to, among other things:
Face recognition, counting people, converting data from analog counters into digital values, human face and silhouette detection, specific object detection such as counting animals, objects, etc., gesture recognition - these are just a few of the sample applications of this hardware, many other ML (Machine Learning) vision projects are available, for example, on the Edge Impulse website, where you can share your projects with other developers.
How to implement SenseCAP A1101?
Ease of deployment of the camera is another advantage of the model. The process of connecting the camera to the LoRaWAN gateway itself is not complicated, and the device will be compatible with equipment from all manufacturers that support LoRaWAN technology. The manufacturer, Seeed Studio, additionally offers its application, which, with the help of an intuitive dashboard, will help us use all the functions of the device and simple implementation. The camera data is also automatically prepared for access via http, mqtt API and is available for integration with existing systems.
Additional capabilities
An extensive package of accessories, such as antennas and mounting brackets, is also available for Seeed studio devices. We also offer many other devices based on LoRaWAN technology that will help turn our space into a smarter one at a low cost. All of this using once put infrastructure with minimal fixed cost.