Intelligent IoT edge solutions using AI, ML, and TinyML on embedded sensor devices connected wirelessly
What's at the heart of our technology solutions?
Our experience and understanding of artificial intelligence, machine learning, and TinyML, in the cloud and on embedded devices tells us that the technology is ripe and ready to be applied at all scales today.
Whether we apply AI and ML models and algorithms on the cloud or on sensors the size of a grain of rice, the MLOps toolchains and MLOps processes required to deliver real world models that deliver constrained human intelligence on some of the world’s most smallest and power efficient embedded devices is at the heart of what we do at Quantuity Analytics.
Our experience applying the latest AI, ML, and TinyML research and solutions onto embedded edge devices to do specific jobs of intelligent sensory perception, in almost any environment, is the unique winning value proposition we deliver to customers in the commercial trucking industry.
We ship trained human-like intelligence in highly specific domains, to perform a specific sensory inference, that allows humans to reduce their cognitive burden or extend their cognition and sensory perception to almost any part of a commercial vehicle while the vehicle is at rest or while moving. Our intelligent IoT embedded sensors using TinyML models deployed on each one allows us to perform real time monitoring of the physical behaviours and physical environment to inform the commercial vehicle owner and operator remotely of its operating condition. We train models to know the difference between normal and abnormal conditions of every deployed sensor, on the sensor device, that will communicate back to the commercial vehicle owner and operator of its operating condition.
With multiple task specific sensors distributed throughout a commercial transport truck monitoring important and critical components and aggregated to a central gateway device connected to the cloud via a 5G modem connected gateway, each and every vehicle with our connected fleet system can be remotely monitored for normal and abnormal conditions of many, and especially, critical systems required to keep the vehicle moving while predicting catastrophic component failure before they occur.
Our systems get better with time and with more quality data used in our MLOps processes to improve the AI, ML, and TinyML models used to make our sensor predictions more accurate and the data more valuable over the lifespan of the equipment and intelligent IoT devices.
All of our systems connect to AWS IoT Cloud for a fully integrated intelligent IoT ecosystem. Customers get access to a custom portal and configurable IoT sensor dashboard where they can track many different sensor parameters of each commercial fleet vehicle in real time, anywhere and everywhere 5G connectivity permits via a mobile cellular network.
The value of 5G connectivity in our systems?
5G connectivity provides data speeds up to 3 GBPS under normal conditions. Recently, lab tested have verified speeds of up to 10 GBPS at 4 km distance!
5G mobile connectivity provides extremely low latency of every connected Quantuity Analytics connected commercial transport vehicle monitoring system that allows real time monitoring of connected IoT sensors to tell the fleet manager what’s happening with every asset, shipped freight, and driver status and safe operation of the commercial vehicle. Similar to robotic cameras that take pictures of speeders and digitally share license plate photos with traffic rule violations, the Quantuity Analytics commercial vehicle transport monitoring system is also similar where 5G connectivity allows for real time event monitoring of fleet assets so that corrective actions can be taken to help keep drivers safe, the vehicle operating optimally, and getting the freight from A to B on time.
Lastly, 5G connectivity’s high data rates can allow for the commercial vehicle operator to enjoy downtime while waiting for loads or when off-duty with built-in WiFi data connectivity. The same AI system used to monitor the vehicle in real time can be also used as in-vehicle entertainment when the vehicle is not moving and parked. This is similar to how Tesla’s self-driving car system can be used to entertain its owners to consume content or even play video games in the vehicle. This is an attractive value proposition to the younger connected driver that would like to make the most of downtime between loads.
Intelligent IoT with TinyML on 5G connected edge devices!
Foundation brakes slack adjuster sensor
We’ve developed a POC/MVP intelligent IoT solution to monitor and measure foundation brake slack adjuster travel distance using a TinyML model deployed onto an intelligent IoT sensor. This system measures the exact travel distance of the brake chamber push rod without interfering or fouling with the operation of the brake system or components.
Fan clutch lifespan sensor
We’ve developed a POC/MVP intelligent IoT solution to monitor and measure the function, performance, and usable life of a fan clutch system and fan clutch engagement relay and solenoid under the hood of a commercial transport truck diesel engine. The fan clutch is a critical part to the safe and continued operation of a commercial tractor to keep the engine and components under the hood cool under load. Failure of this part over the road will render the vehicle inoperable and increase the cost of repair outside of its service cycle. Predicting when this part is near its required repair cycle maximizes the usable life of the part for worry free miles of use and preventing over the road costly service due to catastrophic unplanned failures over the road.
Van trailer door monitor
We’ve developed a POC/MVP intelligent IoT solution to monitor and measure van trailer door operation using wireless RF detection. This allows for deterministic measurement of secured loads where the doors must remain closed until the freight reaches its destination. This is useful for high value freight that need additional security measures to prevent freight theft, tampering, or smuggling in an enclosed van trailer. Our sensors are suitable for barn door van trailers and also for roll-up door trailers using TinyML models to predict their function, operation, and state (open closed / up or down).
Driver safety and compliance AI computer vision system
We’re developing a POC/MVP intelligent IoT AI embedded computer vision system that detects driver behaviour behind the wheel while operating a commercial transport vehicle. Our AI model uses transfer learning and pose estimation to predict safe and risky behaviours of a commercial transport vehicle driver so that corrective enforcement of positive driving and operating behaviours can be reinforced. This system maintains driver privacy by only detecting the behaviours and anonymizing the individuals behind the wheel of their image or likeness. Risky driver behaviours are alerted to dispatch over 5G wireless to centralized dispatch to ensure driver health and safety is addressed. We use voice instructions and notifications to the driver when risky behaviours are detected and also voice recognition when a driver needs information for navigation or system checks. All of the features and functions run on a powerful edge AI embedded system with an AI accelerator built-in for extremely low latency AI and ML inference on an embedded edge computing system. The same system can also be used to provide in-vehicle entertainment for the driver when the vehicle is stationary. WiFi connectivity for the driver is also provided with an installed 5G modem connected to a 5G cellular network.