Getting started with an A.I. App

We all know the famous saying, time is money. The more time you spend on something, the more costly the task becomes.

This is certainly true of the time spent performing inspections, and the resulting workflows.

Take the following hypothetical example

Current maintenance for a 200 Ha (500 acres) solar plant takes at least 1 month for 2–3 people team to perform. A very manual process with limited collaborative data analysis. Such a farm may have some 400,000 panels plus other equipment of which a small percentage would be faulty.

Automating the process and utilizing machine learning algorithms with thermal cameras to conduct the drone inspections can increasing productivity 10x and provide a dynamic digital passport of the farm, ensuring fewer replacements and higher revenue generated. This is just one use case.

Many industries rely upon visual inspections, Oil & Gas, Power, Mining, Construction, Public Safety, Emergency Services, Rail & Road, to name a few.

In every instance, the aim is to take unstructured information (an item is missing, defected, etc.), scribe it into a formal discernible data set for further action.

Industry cherry-picks the most critical of needs where visual inspection is a must-have as they can be costly.

Now picture lowering your operational costs and increasing efficiencies without the overhang of safety concerns by scaling your visual inspection with an A.I app built to run on visual data.

So you think this could be something of interest but need some more concrete examples:

• Automate faulty Solar Farm panel inspections.
• Automate inventory and inspection of Transmission lines
• Get analytics on how productive machinery is on a construction site
• Auto-detect methane plums on Oil & Gas facilities
• Perimeter intrusion detection on a facility

These are just a few examples where cameras (Drones, I.P. cameras, mobile devices) connected to the Unleash live platform can run an algorithm to reduce the need for manual intervention.

We are augmenting the role of your people, from the costly task of field-based inspection to reviewing video clips of identified interest.

Sign up to our newsletter to stay up to date with the latest industry applications of Unleash live

So what does it take to get started

A clearly defined scoped outcome

When using people to conduct inspections, they bring their general intelligence. They are sent to conduct checks for missing category x items, and consequently, they also gather knowledge of category y.

A.I. is narrow intelligence. We are training for a specific outcome of interest. Count, identify, inspect, validate, locate. A clearly defined scope for an A.I. algorithm ensures both high levels of accuracy and the ability to repurpose the model created to different sites and situations.

Visual Training data

Organizations with visual data have the raw ingredients to build the perfect model. One thousand images and 60 minutes of video footage from the items of particular interest is a strong starting point.

For example, a Utility company uses UAVs to conduct inspections. The video and imagery data is captured, and labouriously annotated to describe which items are defective with their geo-location. This is the basis for their custom A.I. model.

The new workflow with Unleash live looks like either of these two scenarios.

Get started now with a free trial

Scenario A: Network connectivity exists on site

• UAV mission planned, airspace clearance obtained and flight is conducted. This flight is supervised from a remote control center
• A.I. app runs on the live video and creates video clippings of the items of interest
• Output video clippings of interest are available in real-time, accessible from their Unleash live account anywhere in the world
• The UAV operator can share a URL of the live stream for inspectors to join and collaborate with others, including the UAV operator

Scenario B: No Network connectivity on site

• UAV mission planned and conducted
• The footage is uploaded at a later date to Unleash live
• A post-process job is completed to analyze the video for relevant insights
• Output video clippings of interest are available for review and action

What does it take to get started

Surprisingly, less than you would imagine.

Organizations often raise costs and complexity to implement as critical concerns for executing A.I in their organization. To overcome the costs associated with A.I., Unleash live runs in the cloud through a SaaS model with service fees based upon the number of connected cameras and hours of analyzed video. To build a custom A.I. model, get in touch with one of our partners or contact our customer success team directly.

If you have an existing UAV program, getting started is fast and straightforward.

Get Started with Unleash live

Visit unleashlive.com to start your free trial or book a time with one of our specialist team members to learn how your business can take advantage of live streaming video and A.I. analytics.