Smart parking solution aims to output real-time parking occupancy information. It helps to reduce parking bay search time, traffic, fuel consumption, and thereby vehicular emissions with increased road safety. A computer vision-based solution using camera video data is most reliable and rational since it allows monitoring the entire open-air parking area at once. A real-time parking solution (cloud-based, server processing, or onboard processing) helps bring the occupancy information to the end-user. It comes with many challenges such as viewing angles, lighting conditions, model optimization, reducing inference time, and many more real-world challenges. Hence, this paper presents a case study on real-time open-air off-street intelligent parking management using a deep neural network. Also, most of the earlier research works focus on day-time data and do not discuss the night data. So, in this work, we perform experiments on realtime 24-hour data from an input camera video source mounted to monitor parking at IIT Hyderabad (IITH) parking lot. Our experiments demonstrate the real-world challenges and can help improve parking performance, deployment at IITH, and relevant parking systems in general. © 2022 IEEE.