The SaaS platform by iGloble Software Solutions for connected vehicles enhances operational efficiency, reduces overheads and predicts a number of scenarios to diminish downtime for fleet owners and increase RUL. Dr. Amit Shekhar, Founder & CEO, iGloble Software Solutions, talks to Sarada Vishnubhatla about what gives iGloble a distinctive edge in predictive and prescriptive analytics.
Headquartered in New Delhi and with a presence in Pune in India, Hong Kong, and the US, iGloble Software Solutions has developed an open architecture SaaS based Connected Platform that offers various modules that enable predictive analytics scripted solutions across multiple verticals. These solutions offer operational and engineering efficiencies, cost reduction, safety and security management. iGloble’s iotaSmart Solutions are driven by artificial neural networks (ANN) as part of Artificial Intelligence (AI) and Machine Learning (ML) by creating a digital twin that can be used by fleet owners and operators, tier-1 manufacturers and OEMs in virtual designing, testing and simulation.
Data derived from various types of IoT devices installed in vehicles is managed, cleansed and transformed into a format that is analyzed using master data management, a technology-enabled discipline that works with business and information technology. The elements or variables are simulated for predictive analytics, and the output pushed onto customized the dashboards for decision making with the goal of lower down time, optimized fuel consumption, mapping and predicting driver behaviour, for enhanced safety and reduced operational costs. The data is also used to calculate the remaining useful life (RUL) of components the vehicle as well as for virtual design and testing.
Dr. Amit Shekhar, Founder & CEO, iGloble Software Solutions, shared says “There are a few companies which talk about artificial neural networking or ANN. Many offer diagnostics but what they actually do is pick up data from the CAN and merely publish it. As an innovative company, what we do is that we also get DTC or Diagnostic Trouble Codes along with engine parameters like RPM, speed, engine load, AT and ECT to offer prediction and prescription. This is where AI comes into the picture and the data can be used to build models as well as offer forecasting. Our domain knowledge helps us to merge modelling and AI into one and give more analytical data information so that our customers are able to reduce cost of operations by between 6 to 10% annually.”
The relevant decisions – that need data on idling fuel consumption, emissions, torque, speed and time, operational costs predictive maintenance, carbon footprint, driving score, risk and safety – are all driven by the ANN. iGloble has over 3000 vehicles using their iotaSmart network currently.
Dr. Shekhar adds: “We are focused on connecting vehicles into a single intelligent network across the globe. What sets us apart from our peers is that we capture data every 1 to 3 seconds unlike most of others who do it every 10 to 15 seconds. Every second of data reading and analysis enables us to offer predictive maintenance and solutions for failures. We also provide early diagnosis of vehicle problems using OBD devices besides the basic solutions such as geo fencing, remote locking and unlocking, engine cut-off, and fuel level measurements.”
iotaSmart Solutions
iGloble offers various unique technological solutions using their proprietary iotaSmart platform.
iFLEET is the fleet management solution that collects data in real time. It transforms the collected data into meaningful information, predictive analytics and insights using ML and AI, enabling fleet owners, operators and drivers to make informed decisions in real time for optimizing operations and reducing costs.
The solution also creates a maintenance calendar in real time, that manager. It shows the vehicle number and what must be replaced or repaired based on predictive analytics of the real-time data, ‘a reinforcement of the prediction offered’.
iPREDICT offers predictive maintenance solutions that have been designed with open APIs. The digital twin solution acquires data, calculates the RUL and predicts failures for brakes, tires and chassis. The ML-powered solution integrates data and domain expertise with driving behaviour, trip information, environment conditions and terrain information resulting in highly accurate failure predictions and avoidance of sudden downtimes.
iSAFE, is a solution that uses computer vision algorithms and pattern recognition in an edge-based camera technology, to identify driver fatigue and distraction, driving performance and pre- and post-accident evidence. The algorithms also aid in handling emergency situations and providing evidence in insurance schemes. This solution is designed to enhance safety of people and cargo.
iGREEN is a solution for monitoring emissions for vehicles and fleets. Using iGloble’s proprietary IoT devices, the solution measures emissions and aims to reduce them by between 8 to and 10 %. It also aids in predictive maintenance and modifying driving behaviour and a reduced carbon footprint.
iDESIGN is a data-led component design solution that creates a digital twin using computer simulation models that are used for feasibility determination and performance testing of a components in a virtual environment. It simulates operating conditions, component parameters and climatic conditions using real data to predict exact wear of the component and design analysis reducing the dependency on tests in the laboratory or on the road. It has helped users reduce production costs by between 5 and 7 %.
iGloble’s solutions enable comparative studies of vehicles for fleet owners based on various parameters.
Dr. Shekhar elaborates with an example: “We have created an advanced driving behaviour monitoring system using ML. This is a learning system which continuously undergoes change in synchronization with our progress, based on the type of the vehicle, driver and terrain. Say a certain driver records a high RPM while driving and we can figure out by the data recorded that he does so at least 45% of the time he drives, and he leans towards hard acceleration 2% of the times and so on. Now, that translates into an uncomfortable riding experiencing for the passengers, which is also risky and potentially damaging to the vehicle, tyres, brakes and increases fuel consumption. In another scenario, if the driver is idling, for say, 30% of the time, it could be due to traffic or he may have left the engine idling while taking a break at a wayside amenity. In a nutshell, these parameters can be captured and the direct impact of high RPM, high acceleration and hard braking can be made clearly evident to the fleet operator. In terms of diagnostics, we can gauge the engine health by tracking parameters including engine coolant temperature or ECT, intake air temperature, fuel consumption, voltage with the help of AI.”
Keeping People and Vehicles Safe
A variant of the iSAFE solution, designed especially for passenger safety, is a five-camera solution to detect five different elements – driver fatigue, traffic signals and violations, passenger monitoring, rear view of the vehicle and safety of the luggage. The system generates alerts and the fleet manager or driver can use to make a decision in case of any incident or a predicted incident.
Dr Shekhar adds: “We are now upgrading the solution by connecting it to the IoT Hub where we can send the information about the vehicle speed, driver drowsiness and the Hub manager can take a call for the next action and steer the vehicle towards safety. This is what we are headed towards. And we have a good accuracy rate of 90 to 92%.”
The technology to enable driver face recognition is available in the camera itself, according to Dr. Shekhar, which iGloble has developed in-house. It helps the fleet managers assign drivers to vehicles and then confirm that the drivers are matched with their vehicles according to the roster. Or else, the vehicle gets switched off instantly by enabling remote engine cut-off.
Growing number of Successful Projects
iGloble has built up a track record of successful development and implementation of a number of projects across various verticals from fleet owners/operators to OEM manufacturers both in India and overseas. A recent one is the successful development and implementation of real time tracking of the vehicles, advance ticketing, and dynamic scheduling of routes for a minibus service in Hong Kong.
Dr. Shekhar adds: “Our endeavour is to impact transportation management and optimization and create solutions for industries that are predictive. We make it efficient by streamlining the process end-to-end using ML and AI. The solution is an amalgamation of vehicle, driver, engine analytics, intelligent decision making, route optimization and advanced booking.”
Closer home, iGloble has had similar success while working with SRTUs, fleet owners and logistics companies. He adds: “Our open architecture allows clients to send data to us and they can get access to their predictive calendar, their driver behaviour, and trip reports. Our platform allows us to work with their existing solution to improve efficiency. For all our clients, we gather data like idling fuel consumption, emissions, torque, speed, time in practice, operational costs, predictive maintenance, carbon footprint, driving score, risk and safety using ANN. Coming to the logistics and supply chain sector in India, it is unorganized even today. Our solutions offer clear visibility of the vehicles, asset utilization, which offers enhanced safety. We believe that visibility if improved by even a 2 to 3% will yield an increase in accuracy and save huge amounts of revenue. If we talk about fleet owners, 80% of them have either less than or about 30 trucks. Those with big fleets of 1000s are better organized as they use smart connected vehicle solutions. We also work with SRTUs for whom we have saved operational costs and other resources.”
High Accuracy Rate
Talking about the accuracy of the ML and AI technologies that iGloble has developed and uses, he shares: “We can easily achieve over 95% accuracy and as we progress, over time, we use what is called reinforced learning. It means if a prediction is inaccurate then we can pull it back into the ML loop to improve accuracy as we go ahead.”
iGloble is well on its way to impacting public transportation around the world through its cutting-edge ‘thinking’ technology.