Page 797 - AI for Good Innovate for Impact
P. 797

AI for Good Innovate for Impact



                   Use Case 4: Low-Cost Decision Making for Group of Vehicles for

               Safe Mobility in V2X                                                                                Transport  4.10: Intelligent













               Country: India

               Organization:  Tata Consultancy Services Limited

               Contact Person(s): Primary contact: Garima Mishra, garima.mishra2@ tcs .com, 9811964569

               Secondary contact: Sai Bhavadeesh Yarlagadda,  saibhavadeesh.yarlagadda@ tcs .com,
               8106263461, Hemant Kumar Rath, hemant.rath@ tcs .com, 9035017741


               1      Use Case Summary Table

                Item              Details

                Category          Low-cost and Safe Transport, V2X, 5G/6G

                Problem           1.  Change in Quality of Service (QoS) drastically within a short period due
                Addressed            to high mobility and fading of the wireless channel,
                                  2.  Pre-programmed ML algorithms,
                                  3.  High cost of on-board data computation

                Key Aspects of  Enhance the safety of vehicles, reduce the cost of expensive on-board data
                Solution          computation and analytics, and predictive QoS for vehicle platooning in
                                  multi-environment scenarios.
                Technology        C-V2X, Predictive-QoS, multi-environment, Explainable AI (XAI), Vehicle
                Keywords          platooning
                Data Availability  Public dataset: 1. Berlin-V2X Dataset[1]
                                  2. TiHAN-V2X Dataset[2]

                Metadata (Type of  Time-series data
                Data)
                Model Training  Data pre-processing (Data cleaning, Feature reduction and selection),
                and Fine-Tuning   Lag-variable optimization, Trained on LightGBM Regressor for real-time
                                  deployment, SHAP analysis for Explainable AI (XAI).

                Testbeds or Pilot  Virtual simulations in controlled environments
                Deployments

                Code repositories  Fraunhofer HHI 2024 [9]
                                  Takalani95 2024 [10]









                                                                                                    761
   792   793   794   795   796   797   798   799   800   801   802