Ingenic T31-based WiFi AI camera development kit promises over a year of battery life

We’ve previously seen WiFi security cameras promising a year of battery life with products like the Eufy EverCam, but the Ingenic T31-based Smart Video Application Development Kit by Innophase may allow for the development of even more power-efficient WiFi AI security cameras with AI processing last can last over one year.

The development kit combines Ingenic T31 MIPS & RISC-V camera SoC with Innophase Talaria TWO INP101x ultra-low-power (57µA @ DTIM10) Wi-Fi & BLE wireless module, that is estimated to last 14.4 months on a 3,000 mAh battery while operating at a 99.3% idle, 0.7% video capture ratio.

Development kit content:

• Board with Ingenic T31 MIPS processor @ 1.5 GHz and RISC-V low-power core, H.265 encoder, 512Mbit or 1Gbit on-chip memory fitted with
  • 2M pixel Full-HD camera @ 30fps
  • USB Debug board
  • Microphone and speaker
• InnoPhase Talaria TWO SDIO Adapter Board
  • Fitted with INP1014 LGA module with 2.4GHz WiFi 4, Bluetooth LE 5.0, PCB antenna
  • Power consumption: 57uA @ 3.3V in Wi-Fi Power Save Mode 802.11b, 1Mbps (DTIM10), 19 uA in deep sleep mode
  • JTAG programming and debug header
• INP3000 programmer board with cables
• 12V power supply

The kit also includes Linux drivers for Ingenic T31 and AI algorithms for human detection, facial detection/recognition, cry detection, vehicle detection, pet detection, as well as quick start & user guides. The MIPS processor and Talaria TWO communicated over a “Dual-Stack” software implementation that shares processing responsibilities between the two chips in order to reduce the system’s current consumption to the lowest level. The devkit also integrates device provisioning over BLE, AWS Cloud and Azure Cloud connectivity, and support for firmware updates over the air (OTA).

Technical details about battery life estimation:

• Video Capture: 1080p @ 25fps, H.265
• 6 video capture events per day
• 10-second video capture length
• Wi-Fi – DTIM10, 3min Keep-Alive
• Active / Idle ratio –  Wi-Fi idle connected: 99.93% per day; Active Video Capture: 0.07% per day
• Battery Consumption Breakdown – Wi-Fi idle connected: 30% per day, active video capture: 70% per day
• Battery: 3,000mAh @ 3.6V battery

Those conditions should result in the kit lasting 1.16 years, or 14.4 months on a charge. WiFi idle power consumption is extremely low, so that’s active video capture that consumes the most, and this scenario corresponds to just 6 captures of 10 seconds per day or a one-minute video which will probably not be realistic in most environments. But at least it should be possible to make AI cameras that last for several months on a charge in most conditions.

The kit is available in limited quantity and access to the InnoPhase SDK requires signing an NDA. More details about the development kit can be found on the product page and the press release.

Thanks to TLS for the tip.

Jean-Luc Aufranc (CNXSoft)

Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.

Support CNX Software! Donate via cryptocurrencies, become a Patron on Patreon, or purchase goods on Amazon or Aliexpress