AN ARTIFICIAL INTELLIGENCE (AI) DRIVEN HYBRID EDGE-CLOUD COMPUTING BASED CONTROLLER FOR LI-ION SMART
Abstract
Proposed work offers a combined edge-cloud computing based scalable deploy ability solution to manage safe as well as efficient oper,ation of high power Li-ion battery based autonomous systems. The minimum viable product (MVP) aims at demonstrating the product with, 4-10distributed devices using local edge computing combined with a centrally secure cloud on Purdue high Performance Computing syst,ems using NVIDIATM GPU computing. --The MVP with focus on satisfying the following objectives:1. Objective 1: Parasitic BMS hardware, integration including edge computing-Questions to be answered are: A. Can parasitic (device agnostic) BMS be integrated with Li-ion, batteries in various devices? B. Can edge computing microprocessor collect data from device operation and then process it for cloud, communication?2. Objective 2: Secure edge-cloud connection- Questions to be answered are: C. Can efficient BMS-cloud encrypted comm,unication be established over all varied devices? D. Can the communication speedup and a balance of computing load be achieved based, on device types between edge and cloud?3. Objective 3: Data science solutions for integration of data in secure cloud- Questions to, be answered are: E. Integration of various databases in providing decision metrices should depend up on operational factors. Can su,ch a combination be effective at managing load and safety? F. Can the data point for load balancing and safety communicated to field, devices be dynamican devolving as a function of operational environment change?
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 01, 2022
- Source ID
- N000142212079
Entities
People
- Vikas Tomar
Organizations
- Office of Naval Research
- Purdue University
- United States Navy