In cities around the world, finding a parking spot is a daily struggle. As traffic grows and urban areas expand, managing parking spaces has become more than just a local issue. It’s now a key part of improving traffic flow, reducing fuel waste, and making city life more efficient. This is where artificial intelligence (AI) comes in, and more specifically, predictive analytics.
Subhasis Kundu, a Senior Principal Solution Architect with over 15 years of experience in enterprise technology, has played a leading role in using AI to make parking smarter. He helped build an advanced smart parking system that uses Microsoft Azure AI and Amazon SageMaker to analyze real-time data and predict parking availability, maintenance needs, and pricing strategies.
The system combines data from IoT sensors, video feeds, and machine learning models. This allows it to guide drivers to open spots, alert operators about equipment issues before they happen, and adjust pricing based on demand. It’s not just about convenience, it also helps reduce traffic and costs, and increase revenue for parking operators.
Sharing his experience, he discussed how he has worked with several Fortune 500 companies and helped deliver large-scale AI solutions with some remarkable results. Amongst major projects was designing a smart parking platform that now supports more than 200,000 parking transactions every day. The system delivers predictions in under half a second and has already shown real-world impact. In one pilot project at an international airport, it improved parking space use by 20% and reduced the time drivers spent searching for a spot by 30%. Maintenance costs also dropped by 15%, thanks to predictive alerts.
What makes the architect’s work stand out is how it connects different systems. Using both Azure and AWS platforms, he built a unified data pipeline that allows faster model training and smoother deployment. Instead of relying on a single cloud provider, this multi-cloud setup takes advantage of each platform’s strengths and avoids vendor lock-in, a challenge many tech teams still face.
But as Kundu shared, there were several hurdles in the process. The biggest one was bringing together different data sources from various vendors. Many parking systems operate in silos, making it hard to get a clear picture. The professional overcame this with his team by creating a single analytics pipeline that integrates all incoming data. Another challenge was making sure the system could handle real-time updates without slowing down, even during busy hours. To solve this, his team used edge computing, which moves some of the processing closer to the devices, speeding up the system and reducing its reliance on the cloud.
Alongside his technical work, Kundu has shared his knowledge through research papers and blogs focused on AI, mobility, and smart infrastructure. He has written about topics like AI for digital twins in facility management, low-power AI for real-time use cases, and how AI can improve public services like parking, transport, and even healthcare. His papers, “Facility Management in the Age of IoT and Digital Twins: AI-Driven Optimization for Smart
Buildings”, “IoT-Enabled Digital Twins for Personalized Healthcare: Real-Time AI Models for Predictive Health and Targeted Treatment,” are few such examples.
Looking to the future, the expert believes that predictive analytics will offer more than just operational improvements. He sees it becoming a new way for cities and private operators to create value from smarter pricing models to partnerships with ride-sharing and logistics companies. He also highlights the growing role of edge AI, which helps systems react faster and protects user data better by keeping it local.
In conclusion, smart parking may seem like a narrow focus, but the lessons learned from it apply broadly. It shows how AI, when used thoughtfully, can solve everyday problems in ways that benefit both people and businesses. The future of mobility isn’t just about new vehicles, it’s about using data and technology to make urban life more seamless, efficient, and responsive.