CXOToday has engaged in an exclusive interview with Mr. Neil Unadkat, Co-founder and CTO, Intangles
- Please tell us about your firm and the services it provides globally.
Intangles Lab started operations in 2016. Working with physics-based analytics and machine learning to simulate the real-world environment into a virtual world, we provide real-time and predictive insights augmented with a large repository of repair strategies and recommendations. Our solutions allow fleet operators to monitor, benchmark and conduct predictive maintenance of assets in order to enhance their overall uptime & profitability.
Our solutions are primarily focused on predictive maintenance around the health of the vehicle, driver behaviour analysis and efficient operations automation. Our powerful predictive models for critical engine functions enable alerts well before a breakdown occurs. These notifications are coupled with appropriate instructions to mitigate significant losses. We have taken the novel approach of creating digital twins of specialised power-train functions such as battery charging, engine cooling, fuel injection, and assisted air intake. Historic and real-time data helps deliver alerts of possible failures, leading to a significant reduction in the on-road breakdown of vehicles, thereby increasing operational hours and lowering maintenance/repair costs. With Intangles, you have the necessary information related to fault codes that may arise in your vehicles, whether minor, major, or critical ones, at your fingertips, and you can look at remedies depending on the severity. Fleet owners benefit from higher vehicle availability, reduced cost of consumables and accident-related liabilities.
- Which technologies have been used in the product, and how are we easing operations & communication in the EV industry?
We have created machine learning workflows that run on algorithms that can interpret the physics behind the estimated range of an EV. Range prediction has always been a major roadblock when it comes to the widespread adoption of EVs, and different vehicles on the same route are found to exhibit a high level of variance in battery discharge rates (2%-15%). In addition to that, the Digital Information System (DIS) estimates for Distance To Empty (DTE) are highly unreliable. As a result, ad hoc charging sessions based on spurious DTE readings lead to schedule disruption. Intangles’ platform provides comprehensive data on the number of charging cycles from the moment our device is installed on the vehicle, as well as backtracked data from the moment the vehicle hits the roads. This is done by taking into account the Battery Management System (BMS) degradation levels over time.
We also provide accurate SOC and DTE predictions considering varying ambient and driving conditions. The data comprises diverse geographies, topologies and ambient conditions across which a representative population of vehicles is tested. The machine learning models that are being deployed are able to perform multiparametric forecasting across different variables. In addition to weather forecasts, the model has been trained to make predictions around motor torque, wheel speed and sunset-sunrise trends, which inﬂuence HVAC and lighting. This multi-parametric approach enables consistently accurate predictions across dynamic ambients, traffic conditions and routes. Our platform is able to generate meaningful insights by taking all these factors into account. Our models are able to reach 94-96% accuracy levels which are miles ahead of the instrument panels of the vehicles.
- Growth of Intangles since inception and expansion plans
Our passion for data sciences and automobile technologies led us to the exploration of On-Board Diagnostics data on passenger vehicles. It was fairly discernible that there was limited scope for predictive health diagnostics on passenger vehicles for good/routine upkeep of personally owned and operated vehicles. This led us to the exploration of data streams on commercial vehicles, including trucks and buses.
With a clear use case in sight, we developed our own hardware interface capable of collecting data from CV (Commercial Vehicle) platforms across OEMs, fuel injection and emissions technologies. This was augmented with a state-of-the-art edge-to-cloud communication backbone and a suite of proprietary algorithms targeted towards predictive health alerts for engine overheating, battery and alternator failures, driver behaviour profiling, fuel pilferage and geospatial intelligence.
We are aiming to deploy our devices across the entire Commercial Vehicle segment. This also involves vigorous revamps in the Electric Vehicle segment through our extensive Ambient Cognitive AI technology that gives you real-world performance numbers. We are helping organisations meet the regulatory emission norms in accordance with CPCB – 4 and IUMPR requirements according to OBD regulations. We are also working towards bringing Over-The-Air (OTA) software updates for the ECUs. We are also helping fleet operators stay ahead of the curve by getting better visibility on complex powertrains and simplified analysis of their fleet’s health and daily operations.
- What is the future of the EV sector from a growth standpoint?
With Artificial Intelligence and Machine Learning taking centre stage across numerous arenas in the automobile industry, the EV sector is the most prominent. It is expected to continue to grow in the future, driven by several factors. Governments around the world are introducing regulations to reduce carbon emissions and promote the use of electric vehicles, which will likely increase the demand for EVs. Advancements in battery technology are expected to continue, which will likely lead to increased range and reduced costs, making EVs more attractive to consumers. As more and more companies enter the EV market, competition is likely to increase, which will likely lead to increased innovation and more affordable prices. Growing awareness of the environmental and economic benefits of EVs is likely to increase consumer demand for these vehicles.
As the number of EVs on the road increases, so will the need for charging infrastructure. Governments and private companies are investing in the development of charging networks, which will make it easier for EV owners to charge their vehicles. With the growth of renewable energy sources like solar and wind power, the carbon footprint of EVs is likely to decrease, making them an even more attractive option for consumers. As ride-sharing and car-sharing services continue to grow, the demand for electric vehicles is also expected to increase. Based on these trends, it is expected that the EV market will continue to grow at a significant pace over the next decade, with the global EV market size projected to reach $777.7 billion by 2027, registering a CAGR of 25.5% during the forecast period 2020-2027.
- Kindly brief us about Intangles’ journey so far
Today we have onboarded 7 OEMs in the 11 countries where we are now operating. Furthermore, the platform already has over 8,000 fleet operators. We enrol around 800 fleet operators every month and collect an astounding 5 billion sensory data points per day. We estimate 5x growth in FY’23, with some of the top brands in mobility already signed up as customers. Our target clients are leading OEMs in the heavy, light and medium commercial vehicle segments, along with medium and large-scale fleet operators. Our main focus is on the transportation, electric vehicle, construction equipment, power gensets, mining and hybrid powertrain segments. While we are cementing our position in the Indian mobility ecosystem, the prospect of new opportunities in North America, Europe, Australia, and APAC is highly promising. Our remarkable development and expansion story exemplifies the game-changing potential of Predictive Analytics enabled by Digital Twin technology. We are on the verge of ushering in a new age of automotive performance and safety innovation in India and throughout the world. The remarkable progress that we have experienced as Digital Twin in Mobility pioneers is not surprising.
Source : cxotoday