What motivated your organization to participate in this year’s Global AI Awards?

Participating offered an opportunity to receive independent industry recognition for the progress we have made in bringing practical edge AI to real-world devices. AkidaTag demonstrates what the market is increasingly asking for: always-on, private, battery-powered intelligence that can operate locally without depending on continuous cloud connectivity.

Could you give us an overview of the AI solution or breakthrough you submitted for consideration?

BrainChip’s AkidaTag technology is a compact, battery-powered edge AI reference platform designed to demonstrate always-on intelligence directly on-device. Combining the AKD1500 neuromorphic processor with sensing, connectivity and an embedded MCU, it supports applications such as industrial vibration anomaly detection and keyword spotting, while providing a flexible platform for additional consumer, wearable and sensor-based use cases.

How did your team collaborate to develop and refine this AI innovation?

Our team developed AkidaTag as an integrated reference platform rather than simply adapting a cloud-oriented AI solution for a smaller device. The design combines BrainChip’s AKD1500 edge AI processor with a Nordic nRF5340 dual-core MCU, onboard sensing, battery management and wireless connectivity. Just as importantly, the team brought together the hardware, embedded firmware, AI models, MetaTF development tools, Bluetooth connectivity and mobile-device workflow into a reproducible platform that customers and partners can use as a starting point for their own edge AI products.

Were there any notable challenges during the development of this AI solution, and how did you overcome them?

A major challenge in edge AI is reducing the integration effort required to move from an AI demonstration to a practical embedded system. We addressed this by approaching AkidaTag as a complete reference platform rather than as a standalone AI component. The system integrates sensing, local AI processing, embedded control, wireless connectivity, power management and the software workflow needed to deploy and update the device. Local processing also means raw sensor data does not need to be continuously sent to the cloud, which can simplify privacy and data-governance considerations. By providing a reproducible hardware and software starting point, AkidaTag helps reduce the engineering friction between evaluation and product development.

How does your organization nurture a culture that drives continuous AI innovation?

BrainChip fosters innovation by focusing on technology that continuously improves even after it ships. Rather than relying on static models or continuous cloud retraining, we emphasize silicon-level innovations. Our hardware-native 1-bit Edge Learning layer allows devices to adapt and personalize to individual users or machines in the field. We also design flexible systems capable of running multiple models simultaneously (such as parallel voice wake-up and activity classification) within constrained battery budgets.

What advice would you offer to teams or companies aiming to make meaningful contributions in the AI space?

Focus on real-world utility, ease of entry, and solving actual market problems rather than building more AI that lives in the cloud. The fastest way to grow the edge AI market is to remove integration friction. Delivering a complete, accessible solution, where hardware, software, and privacy are seamlessly integrated, allows partners to build and deploy meaningful products much faster. 

What are your organization’s long-term goals in AI, and how do you plan to advance the field moving forward?

BrainChip’s long-term goal is to advance the edge AI ecosystem by delivering ultra-low power, event-based neuromorphic processing. Alongside our silicon, IP and software technologies, we see reference platforms such as AkidaTag as an important way to make that technology easier to evaluate, integrate and productize. We aim to compress development timelines for our partners across industrial, consumer, aerospace, defense, healthcare, and wearable markets, making always-on local intelligence a standard feature in everyday devices.

Are there any emerging AI technologies or trends your team is particularly excited about right now?

We are excited about on-device generative AI. In fact, this is the area we are focusing on next: running a complete voice-to-voice pipeline locally, speech recognition through language understanding through speech synthesis, so a device can hold a conversation without a prompt ever leaving it. Alongside that, we are focused on time-series workloads such as streaming audio, vibration and radar, where compute stays constant per sample instead of buffering frames. Those two directions belong together, since the same efficiency that handles a continuous sensor stream is what makes a generative pipeline fit an edge power budget.

To dive deeper into BrainChip's award-winning work, visit their website at https://brainchip.com/akida-tag/