SenseCAP A1101 - LoRaWAN Vision AI Sensor is a TinyML Edge AI enabled smart image sensor. It supports a variety of AI models such as image recognition, people counting, target detection, meter recoignition etc. It also supports training models with TensorFlow Lite.

SenseCAP A1101 - LoRaWAN Vision AI Sensor is an image recognition AI sensor designed for developers. It will be your choice to open the door to the new TinyML world if you are:

- a machine learning developer who wants to develop edge AI and vision recognition applications and is finding a low-cost and reliable hardware platform to get started; or

- a software/hardware engineer who is keen to learn about TinyML, the model training process, and the algorism; or

- an enthusiast simply wants to experience the AI machine learning product without writing a line of code; or
- a SenseCAP user who wants to add a vision sensor to meet a specific application need; or

- a solution integrator or business developer who has specific AI application demands that need to deploy and scale in a vertical industry using a low-cost LoRaWAN network;

SenseCAP A1101 - LoRaWAN Vision AI Sensor combines TinyML AI technology and LoRaWAN long-range transmission to enable a low-power, high-performance AI device solution for both indoor and outdoor use.

This sensor features Himax's high-performance, low-power AI vision solution which supports the Google TensorFlow Lite framework and multiple TinyML AI platforms.

Features :

Ultra-Low Power and Powerful Himax Camara: 400Mhz DSP, Maximum camera frame rate 640*480*VGA 60 FPS, Local inferencing
Low Power and Long Range Transmission: Down to 2.3uWh sleep mode power consumption, powered by Wio-E5 LoRaWAN Modules, transfers data up to miles
High Data Security by Edge Computing: Local image inferencing and transfers the final result data to the Cloud, suitable for applications that require limited data transmission and high data privacy.
Easy to Visualize Data: Few clicks to display and manage data via SenseCAP Mate App and SenseCAP Dashboard, wide compatibility with other third-party tools
High Industrial Protection Grade: -40 ~ 85℃ operating temperature and IP66 rating, suitable for indoor and outdoor deployment
Easy for Scalable Deployment: 1min to add and configure the device by scanning the device QR code, low LoRaWAN network cost and maintenance cost ensure business scalability.

What SenseCAP A1101 could do

SenseCAP A1101 is a product that needs developers' secondary development in order to deploy in a real-world application, and it enables users:

1) to test SenseCAP A1101 with the two demo models provided in the product (people detection, and meter recognition) and experience Vision AI application;

2) to collect images using SenseCAP A1101 and generate a dataset through an annotate platform (e.g.: RoboFlow) ;

3) to train and generate their own models through third-party AI training tools and deploy them to SenseCAP A1101;

4) to cooperate with an AI industry partner on dataset generation, model training, algorism developing, etc. to get customized and deployable models for specific application usages and deploy them in the SenseCAP A1101

AI Model Training Process

With operating temperature from -40 to 85℃ and IP66 waterproof grade, it is suitable for outdoor and harsh environments IoT applications.

Meanwhile, we provide the SenseCAP all-in-one solution for this device which supports the rapid configuration of the Bluetooth APP, fast connection to gateways, and Cloud services.

Travel Data through SenseCAP Cloud for Free

With SenseCAP Mate App and SenseCAP Cloud, it is very easy to get and utilize data with few clicks, and they also support wide compatibility with other third-party tools.

Through the plain and easy tutorials, you can connect the sensors through the SenseCAP Cloud and travel the data by LoRa® within 10 minutes

TinyML Demo Models Provided for Quick Test

Embedded Human Body Detection and Meter Recognition TinyML model demos for a quick test and product experience.

Human Body Detection Demo

Meter Recognition Demo

Applications :

FAQs

What is the detailed specification of the camera in SenseCAP A1101?

HiMax HM0360 Sensor; Lens size 1/6"; 3P+1R Lens Construction; F/No 2.4±5%; Full-color 30M pixel; Max Frame rate 640*480 VGA 60 FPS; Optical FOV 82º; EFL 1.98mm

What is the ideal camera setup for image collection?

For a 30M pixel camera, it is suggested to collect images within 1m-5m to get the ideal image effect. Please make sure to face the camera straight towards the objects and have sufficient illumination (no reflect light, no overexposure), and avoid large movement and vibration when collecting.

How far could the SenseCAP Vision AI camera detect an object and have good results?

Based on our test for the people detection model, we could reach results of 70% degree of confidence within 1m-5m.

Is it possible to change the camera or have a customized version SenseCAP Vision AI sensor?

Yes, and if you have large quantity demands,

How long does the battery last?

It will highly depend on the image collection and data upload frequency. Based on our test, if the A1101 collects image and upload per minute, the battery could last for 9.5 months.

Is the battery rechargeable? Does it support solar energy charging?

No. If you have large quantity demands on the customized version,

What could I do to improve and have a better result?

The AI result depends on many factors, including the quality of the images, the accuracy of the annotation, the way to generate the dataset, the model training process, the algorism used, etc.

The one you could improve with SenseCAP A1101 Sensor is to directly use SenseCAP A1101 camera to collect the image instead of using the existing images shot by other devices or using the existing dataset purchased from the third party. Especially when you are having a relatively small dataset, it will help a lot to improve the result.

And check your image collections to make sure they are having the right exposure (avoid insufficient or too strong light), right angle towards the objects, etc.

What is the difference between Grove Vision AI and SenseCAP A1101?

SenseCAP A1101 - LoRaWAN Vision AI Sensor is an industrial-level SenseCAP sensor that has an IP66 grade industry-level enclosure and battery, which has a Wio-E5 LoRaWAN module and MCU embedded. Benefiting from the reliable SenseCAP sensor platform, it is very convenient to add SenseCAP A1101 to the SenseCAP Mate App and Dashboard and deploy the application when the model training is completed.

Grove Vision AI Module, which is provided in the SenseCAP K1100 - The Sensor Prototype Kit and needs to be developed with a mainboard with MCU (e.g.: Wio-Terminal or Arduino Uno). Grove Wio-E5 module is required to connect to the mainboard if LoRaWAN transmission is needed; If the low power consumption feature is not the main consideration, it could also connect to the Cloud via WiFi. (Click here to learn more)

Will there be more AI-related sensors developed by SenseCAP?

Now we only have SenseCAP A1101 Vision AI Sensor on the list. The next would be SenseCAP A1102 - LoRaWAN Sound/Vibration AI Sensor, but at the moment it is still under development. Please stay tuned.

ECCN/HTS