Management Perspectives

Just as electricity transformed lives upon its discovery, AI is changing and is continuing to change our lives completely. Andrew Ng, co-founder, Coursera and Chief Scientist, Baidu Research, says, “AI is the new electricity.” Google, Netflix, face detection, predictive search, Google translate, Alexa, recommendations, maps, self-driving cars, to name a few, all use some form of AI to make our lives better. In simple sense, AI has already permeated our day-to day activities. Data is the most fundamental input for building AI systems and its implementation is growing exponentially.

AI with its applications in the automotive sector is seen as a disruptive technology. No doubt that by the end of 2030, we can expect about 95 to 98 per cent of the new vehicles be driven by AI technology (FutureBridge Analysis and Insights 2020).

The hardware value proposition will come down; software and content value will rise, resulting in autonomous vehicles being fully run by software value propositions. The AI disruption in automobiles runs in ‘6 Levels’ as adopted by the US Department of Transportation (Synopsis 2020 explaining the 6 Levels of Vehicle Autonomy). At Level 0- the driver fully owns and controls the vehicle, but from Level 1- AI provides some assistance; Level 2- partial autonomy (you can take your hands off); Level 3- eyes off; Level 4- turn off your mind; and at Level 5- its only the passengers and the ‘Driver Off’ mode. There is transfer of responsibility from human to machine. Now, major automobile industries are exploring Driver State Monitoring (DSM) system which monitors the state of the driver (news provided by Technavio, August 9, 2021). The driver needs to pay attention to the road and should be able to take control of the car when needed. Computer vision and deep learning image processing techniques are applied in DSM to identify driver attentiveness, driver analytics based on age, gender, experience, send locations, and to activate autonomous driving.

For example, if sleep is detected, coffee break will be advised. AI is used for anomaly detection for any assembly line defects using 360 degree camera rotation (Chui et al., 2018) to detect each and every parts and also enable supply chain design from the smallest screws to each and every unique parts in a cars. AI is used to forecast demand— time series-based deep learning methods, lead time and travel time designs, predict and manage inventory, and matches demand and supply using AI based tools.

When it comes to sales of automobiles, targeted marketing using AI can run marketing campaigns for customised segments using customer profiling (Huang & Rust, 2021) in the form of clusters based on specific characteristics. This takes into consideration that ‘one size does not fit all’. The pulse of customers is collected by measuring satisfaction at touch points through AI-powered sentiment analysis and by natural language processing using BERT. Again, AI-powered insurance claims will automate insurance products as AI will analyse driving patterns of drivers and assists in modifying the insurance premiums. For risky drivers, insurance premium will be higher and vice versa thus promoting safe driving. Recently, face mask detection also came in the purview of AI in automobiles.

AI is transforming life beyond the expectations, but still feels as a long way to go. The ‘trolley problem’ is still not solved. The ethics in AI-powered decision making is still a question. We are looking for an unbiased, ethical AI solution in all arenas accepting the fact that ‘with great power comes great responsibilities’.