The IoT Ecosystem

The IoT Ecosystem

The differentiating factor of IOT is that it is not an isolated technology, rather a complete ecosystem. This is the reason IoT took it’s time to become mainstream and it is still evolving. However from our point of view, the full potential IoT can be reaped when the following elements come together –

Sensors and Actuators

Sensors and Actuators are the front face of the IoT ecosystem. Sensors collects inputs and actuators provide the actionable outputs. Sensors act as the Sensory system, collecting data about temperature, moisture level, light reading, gas, distance, humidity, motion, smoke, proximity and many other variables. Based on the input data, the actuators perform the desired actions like sound an alarm, open a gate, turn on the sprinklers, turn on/off the lights, alert the security, send a text message and many more. PWM is used to control the flow and degree and Relays are used as switch to turn things on or off. The possible combinations are almost unlimited.

We are currently working with more than 110 sensors in 36 categories. These sensors are capable of measuring almost all types of environment variables and give you an accurate reading of almost anything you want to measure. With this type of information, we are in a unique position to determine what to do with our actuators. We will provide some example in the use case section.

Gateways

Sensors collecting data and actuators doing required tasks are not new. However, the Gateways connecting multiple different types of sensors and actuators is what makes IoT interesting. Nokia was connecting people but our IoT Gateways connect “things”. We have different gateway to connect the sensors and actuators. We have basic gateways (iG 100 and iG 200) for home and small business, and large, high capacity gateways (iG 500) for industrial use.

Our inexpensive iG 100 product connects 4 analogue and 12 digital sensors with 4 PWM output and 4 relay output. It provides on board CANBUS and MODBUS protocol support with Ethernet and Wi-Fi capability. The system is loaded with onchip MQTT server which you can stream data to the cloud such as AWS or you can host your own IoT event processing and decision support system.

iG 200 offers on board FPGA and multicore CPU for larger number of sensors data aggregation and send stream to your internet hosted server in gigabit speed.

IoT gateways will be the first phenomena of the IOT revolution. It is very similar to the concept of “Interchangeable Parts” that kicks the industrial revolution. Currently you need to use separate methods and channels for connecting different types of sensors. Sensors can be Analog or Digital, can have different connection options (RS485, RJ45). Without a proper Gateway, it is impossible to connect and manage all of them. Our gateways have all kinds of connectivity options to connect multiple different sensors and actuators.

Aggregators

For homes and small businesses, the basic gateways are sufficient. However, in case of large installations in manufacturing and in mining, we need a way to connect multiple gateways that connect multiple sensors. It also have to be powerful enough to control such a large infrastructure.

aG 1000 is the high performance computing at the edge with 4 processing module – either all four can be GPU or Xilinx Ultrascale + MPSOCs. You have power of providing real time video analytics or running machine learning at the edge. You can run most of the framework at the edge. We support NVidia TX2 GPU system for your advanced tensor flow and other frameworks.

Analytics

Data is useless unless it is analyzed and meaningful insight is extracted from seemingly unrelated chaos. As the sensors are collecting a massive amount of data, we need to have the computing power to process them either in the Gateway / Aggregators or in the Cloud. Our upcoming Aggregator aG 1000 have the NVidia TX2 GPU have massive processing power to analyze even live video feeds. You can host your own IoT event processing and decision support system or use the builtin MQTT server which streams data to the cloud such as AWS. All our Gateways and Aggregators have Wi-Fi and LAN connectivity to transfer input data to the cloud.

Once the data in in the local IoT Analytics server or in the cloud, you can utilize standard analytical tools and methods (Like R) or specialized Big Data tool like Hadoop, MapD to convert the data to meaningful information.

Machine Learning

The final component of our IoT ecosystem is Machine Learning and Artificial Intelligence which process the enormous amount of data. Human intelligence is still irreplaceable but we can use machines to do repetitive works. And if they can learn to do their job better, it’s a massive relieve for us. We are working on Machine Learning and Artificial Intelligence for almost a decade to know a trick or two. We can take any situation and install intelligence in any system.

Closing Thoughts

IoT is surrounded by a lot of hope, confusion, and dreams. Similar concepts exist before bearing many different names but only recently the full potential is beginning to emerge clearly. Proliferation of data collecting options, improvement in processing power, miniaturization, and maturity of standards all contributed to the massive expectation from IoT. And we are a believer. We started working on IoT as telemetry system back in 2006, and is ready to bring real solutions to you out of the all the noises in the industry.

Quick Links