Addionics reveals battery architecture for always-on AI

Periodic rest is assumed in the battery design; always-on autonomous systems cannot accommodate this, and this gap is Addionics’ thesis Addionics introduced Autonomous Architecture, a redesigned battery cell structure for systems that need to operate continuously without scheduled rest periods. The company says the technology improves current distribution, thermal management and material usage within individual cells, targeting autonomous vehicles, robots, drones and aerospace applications. Traditional battery design assumes periodic outages during which cells can recover from the stress of high demand cycles. As robotaxis, autonomous vehicles, and humanoid robots approach continuous operation, this assumption becomes a limitation: under sustained load, degradation accelerates, available energy decreases, and operating life shortens. Addionics addresses the replacement of isolated electrode layers within each cell with an integrated electrochemical structure. The company says this reduces local stress and improves current and heat distribution throughout the cell, enabling higher sustainable performance in existing and emerging battery chemistries. The approach is compatible with existing manufacturing processes and targets autonomous mobility, robots, satellites and defense systems where operational continuity is a primary requirement. Addionics Chief Executive Officer and Founder Dr. Moshiel Biton said the company’s aim was to meet the demands of machines built to operate around the clock “without any meaningful improvement” and argued that the batteries that make the EV era possible were never designed for always-on conditions. He described Autonomous Architecture as “a new foundation for batteries always built for the open world.” Addionics operates what it describes as the world’s first commercial manufacturing facility for three-dimensional porous current collectors, working with partners in the automotive, aerospace, energy storage and defense fields.