B23 is set to showcase its ruggedized Artificial Intelligence (“AI”) Edge Computing Pod at the U.S. Special Operations Command (USSOCOM) Technical Experimentation 2019 (TE 19-3) this upcoming week July 15th – July 19th, 2019 at the Airfield at Texas A&M Rellis Campus in College Station, TX.
The B23 AI Edge Computing Pod will allow the military to deploy AI-enabled hardware in austere environments to process a variety of sensor data using sophisticated machine learning algorithms.
This capability will also help commercial corporations including retail stores and quick service restaurants (“QSR’s”) perform AI-enabled analytics with data sourced from a variety of Internet-of-Things (“IoT”) devices directly in remote offices and store locations.
B23 AI Edge Computing Technical Specifications:
- Up to 3 8th Gen Intel Core processors
- Up to 192 GB of RAM
- Up to 10TB of local storage
- Kubernetes v1.15
- Docker v18.09
- Up to 3 Nvidia GeForce 1050 GPU
- Up to 6 Google Coral Tensor Processing Units (“TPU’s”)
- Optional FPV 5.8GHz NTSC video receiver with support for 5 bands 40 channels
- TensorFlow r1.13
“Disconnected and low bandwidth comms are the norm when it comes to operating in adverse environments, we [B23] solved the communication problem between Edge Computing and the Cloud and now provide that capability to all our customers’” said Brad Kolarov, B23’s Managing Partner and former Navy SEAL.
“We've optimized AI machine learning algorithms for smaller than usual form factor devices, while also ensuring that we are using the most secure open source software including Kubernetes. This allows us to effectively analyze data using AI, and that we are primed and ready for production for both the Tactical Edge and Cloud,” said Mark Bittmann, B23’s Chief Data Scientist.