F.A.Q.
Product Information
What is Computer Vision?
Computer Vision is a field of artificial intelligence where computers are trained to interpret and understand the visual world, processing images or videos for decision-making.
What are the primary uses of Deepomatic's platform?
Deepomatic’s First Time Right Automation platform leverages computer vision to automate field operations quality control. By analyzing photos of infrastructure assets taken by field workers in real time and giving them live feedback on the quality of their work, we help them do their job right the first time. It allows companies to operate infrastructures more efficiently, streamline their back-office processes, and perform efficient predictive maintenance.
What are the main industries Deepomatic covers?
Deepomatic offers solutions for companies that operate essential infrastructures like telecommunications (Fiber Optic Roll-Out, Fiber Optic Installation), Utilities (Smart Meter Installation), and Emobility (EV Charging Stations Deployment). Our computer vision platform has also served customers in the pharmaceutical and catering industries who need to automate visual-based processes.
Is your solution off the shelf, or do I need to do some adaptations?
With a decade of experience in computer vision project deployment across several industries, we can leverage off-the-shelf models and speed up the implementation process. For bespoke requirements in terms of use case, equipment, or control point, we develop tailored algorithms with new datasets provided by the customer.
What are the benefits of your solution?
The top benefits of Deepomatic’s First Time Right Automation platform include:
- Reduction in truck roll: by reducing the number of errors in operations, companies avoid planning costly revisits to correct them.
- Reduction in Quality Control costs: by using AI in 100% of their operations, our customers reduce the number of field audits and save resources on back-office quality control.
- Sticking to the deployment calendar: First Time Right Automation accelerates the pace of operations by shortening the quality control feedback loop and reducing revisits that usually delay deployment calendars.
To dig deeper into all our platform benefits, read the Total Economic Impact™ of Deepomatic, powered by Forrester.
What if my field operators try to abuse the system and upload duplicate photos?
Our system effectively prevents the abuse of duplicate photo uploads by field operators through several measures:
- Real-Time Detection: It immediately identifies if the same photo is uploaded more than once for a single work order.
- Batch Analysis: We also check large data sets, ensuring no duplication across multiple work orders.
- Advanced Fraud Detection: The system can recognize if a photo is taken of another smartphone screen or a printed image, preventing misuse.
- Direct Upload Requirement: It mandates that photos be uploaded directly from the field worker’s smartphone camera, blocking access to previously stored images in the smartphone library.
Implementation and Timeline
How long does it take to have a PoC ready to go?
Developing a PoC typically takes between 4 to 6 weeks. This timeframe includes the initial perimeter setup, configuration, and testing phases.
Contributing according to our impact
Deepomatic’s platform can be live within three months if our clients follow our recommended integration journey, from data collection to AI training, solution integration, and change management.
For more information, read our Guide Book explaining how to ‘Deploy your computer vision platform and generate value within 90 days”.
What is the average number of photos necessary per work order to ensure quality on the field?
No training is necessary to verify photo quality (brightness, clarity, etc.). However, checking that photos contain the expected elements requires at least 100 photos of compliant examples. Finally, a few hundred photos are necessary to train the AI to verify field staff work conformity.
Technical Details
Once the solution is in production, how is it updated?
The solution may need to evolve. For example, it must be updated if a new photo needs to be analyzed. These evolutions are directly managed within our platform thanks to our versioning system.
Do you provide SSO capabilities?
Our SSO feature works with different Identity Providers, including Google and Azure Active Directory.
What type of API standards do you use?
We provide a REST API to all our customers.
Does your solution work with no internet connection?
It is necessary to have an internet connection so that Deepomatic’s AI can give feedback to the technicians. However, field technicians can take pictures offline and upload them whenever their connection is back.
Data Handling
What are the different types of data that Deepomatic can analyze?
We can analyze three types of data: photo, PDF, and video.
Where is the data stored?
We offer two storage regions on the Google Cloud Platform (GCP):
- Europe: Data storage and processing occur in both Belgium and the Netherlands.
- North and South America: Data is stored and processed in the United States.
Clients are typically assigned to the nearest region to minimize latency.
How does Deepomatic address ethical considerations in AI development and deployment?
Deepomatic is committed to ethical AI development and deployment, ensuring that personal data is never included in our training sets. Our AI systems, deployed in the field, are activated by technicians to aid in guidance and self-control rather than for monitoring purposes. Additionally, we strictly adhere to a policy against using our AI to track or recognize individuals, prioritizing privacy and ethical use of technology.