Ganesh Bhagwat

The existing multi-layer tree is of dynamic linked list type which has many limitations and is complicated due to the pointer-node structure. Static array representation gives more flexibility in programming of algorithms and operations like insertion, deletion, and search. It also reduces the storage space. This paper presents a new method for representing multi-layer forest data structure in array format. It also explains various tree operations, unique data compression algorithm and migration algorithm between traditional approach and the proposed data structure. Most of the fundamental algorithms like those from artificial intelligence that employ decision trees are based on trees/forest data structure. The current paper brings a completely new idea in the representation of these data structures without employing recursion and targeting memory optimizations with reduced code complexities. The applications of forest data structures are many and span over various interdisciplinary areas of Engineering, Medicine, Aviation, Locomotive, Marine, etc. The proposed novel approach not just introduces a new method to look at the tree data structure but also provides the flexibility to adapt to the existing methods as per the user needs. A few such applications in Simulink debugging and the Forest visualisation have been discussed in detail in this paper.

Maintenance of road infrastructure has gathered a lot of attention in recent days. Bad road conditions often cause damages to vehicles and passengers that cannot be neglected. Some of the major factors that cause damage are potholes, speed-breakers, uneven patches etc. The presence of such abnormalities also affects ride quality and safety of passengers. The paper describes a low-cost method that employs a smartphone and an OBD-II module to detect and locate potholes on roads. It utilizes vision data, sensor data and OBD data to create and validate triggers caused by potholes. The image-triggered and data-triggered methods proposed recognize potholes on roads and create a trigger through image/video processing and data processing respectively. The results obtained from each method is cross validated by the other method that help in evaluating ride quality based on road conditions.