Edge AI/ML
Within PodView is a powerful NVIDIA edge processor which hosts the PodView operating system (AdvISE) and onboard AI/ML. The AI/ML is used to autonomously detect, classify and track a diverse range of targets (ISTAR) in real-time delivering accurate, localised intelligence, enabling immediate analysis and decision making.
AI/ML agnostic
The period of relevance of today’s AI/ML technology is very short, sometimes measured in weeks. The AI/ML is interchangeable within PodView, allowing the rapid update of this fast moving technology. Podview allows the selection of AI/ML models and associated training data from multiple vendors or you can bring your own.
Seamless Integration
PodView can be used on its own using the AdvISE web based user interface or as part of a broader solution for example integrated with C² or Battle Management Solutions. PodView includes dedicated SDK and APIs and is also fully compliant with standards such as SAPIENT, ONVIF and TAK. Communication with PodView is via IP allowing the use of a broad range of networks.
24/7 Operation
A low SWaP footprint and remote management capabilities enables PodView to deliver persistent autonomous surveillance and all weather operation for extended periods. When battery operated, PodView’s sophisticated power management extends mission duration.
Military pedigree
PodView boasts a strong military pedigree – built to stringent defence standards, and first adopted by defence users, reliability and robustness are designed in. Leveraging Commercial Off-the-Shelf (COTS) sensor technology, PodView delivers cutting edge performance with adaptability and exceptional value for money across a wide range of applications and markets.
Flexible Deployment
PodView is designed to withstand harsh environments with an aluminium housing, marine grade finish and toughened hydrophobic windows. PodView is deployable on the ground, on fixed structures, on manned or autonomous vehicles, across diverse environments, supporting covert and overt operations.
Stabilisation
Advanced IMU stabilisation technology minimises the impact of vibrations and sudden movements. By fusing both mechanical and electronic stabilisation with AI object tracking PodView ensures clear, stable imagery no matter the environment.
Agility
When the consumer of video data moves from a human operator to AI/ML a paradigm shift is met. AI/ML undertakes image assessment on milliseconds rather than the seconds human users need for the same task. Agility is the ability of the motion platform to take advantage of the speed benefits of AI/ML, allowing the sensors to react at machine speed scanning for new threats or tracking fast moving and evasive targets.
Low latency
PodView has been designed from the ground up to minimise latency and shorten the OODA loop. The revolutionary architecture provides RAW imagery from the head to the data processor in the base with no compression latency, allowing PodView to take full advantage of high-resolution uncompressed data whilst delivering high performance agility.
Low SWaP-C
PodView’s underlying direct-drive motion technology, use of lightweight materials and sophisticated power management delivers class leading SWaP characteristics while retaining the highest performance, environmental and low cost. Using COTS sensors further enhances cost effectiveness.
Information Centic
PodView provides human users with information not data reducing the cognitive burden. PodView ingests data from its multi-spectral sensor payload including optical, thermal, location and ranging, processing the various data streams into actionable alerts containing clear concise information. Whilst focused on the delivery of digital information PodView can also deliver supporting still imagery and full motion video when required.
Sensor agnostic
PodView2500 provides a preconfigured sensor payload of a 4MP visual camera with optional LWIR Thermal Imager and LRF. PodView’s architecture allows Overview to rapidly configure the sensor payload allowing PodView to meet a wide variety of missions. AdvISE allows PodView to integrate with external Specialist Sensor Modules (SSM). Mission specific sensor payloads can be integrated as required.
Technical
Pan/Tilt speed: 0.05° to 720°/s
Pan/Tilt range: 360° with continuous rotation
Edge Processor: NVIDIA Jetson AGX
Sensors: EO (4MP, 30x), LWIR, LRF, GPS, IMU, Environmental
Tours: Unlimited preset tours
Preset Positions: Unlimited quantity with repeatable accuracy better than 0.01°
Digital I/O: 2 inputs and 4 outputs
Onboard data storage: 2TB
IP Rating: IP68
Stabilisation: Mechanical direct drive and electronic image stabilisation
Controls: AdvISE, ONVIF profile S, SAPIENT, TAK
Video Streams: H.264 with optional H.265 software load
How this works
PodView detects movement, prompting AI classification of the object. If it is of interest, PodView continues to track the object, whilst sending an alert to personnel.
AI Partners
Overview collaborates with diverse range of AI providers, integrating their technology into our Remote Autonomous Sensor Systems. Through combining our expertise, we can offer our customers optimal solutions for their needs.
In stark contrast to the long life of a Sensor System, the period of relevance of AI/ML models ranges from days to months. PodView allows users the flexibility to pick and choose the AI/ML model and continually update and enhance capability. AdvISE allows users to swap AI/ML models directly in the field, with minimal downtime, to keep up with dynamic and ever-changing mission profiles.
The ability to change AI/ML software loads on the fly means that PodView offers a multi-purpose solution.
Supplemental Sensor Modules
Podview utilises COTS sensors integrated within the device. Supplementing these native sensors, PodView supports third party SSMs (Supplemental Sensor Modules) which can be controlled by, and whose output data can be ingested into, the PodView edge processing capability. SSMs can be mounted remote to, or onboard, PodView. Example SSMs include radars, ESMs, magnetic, vibration sensors and more.
