Customization for Safety AI

Updated December 15, 2025 22:14

Manage Safety References
Manage Risk Filtering

Availability

Customization for Safety AI is available to project owners and project coordinators with a project running Safety AI.

How to manage Safety AI customization

Follow these steps to access and manage the Safety AI customization modal:

Navigate to the DroneDeploy web app and open one of your projects running Safety AI.
In the top-right corner, click the gear icon to open the project settings in the left side panel.
Beneath the project name and project type, a section is titled "Safety AI".
Click the "Manage Safety AI Customization" button to open the customization modal.

Manage Safety References

Being able to cite a specific safety standard is a core function of Safety AI. The referenced safety standard should align with the standard your company uses daily.

In the customization modal, the default safety standard is U.S. Occupational Safety and Health Administration.
Click the dropdown menu to browse the available safety standards, which include:
- United States of America – OSHA
- Australia – SafeWork Australia
- Canada – Canadian Centre for Occupational Health and Safety
- California – CalOSHA
- Ireland – Health and Safety Authority
- Japan – Industrial Safety and Health Act
- South Korea – Safety and Health Rules
- United Kingdom – Health and Safety Executive

Don’t see the safety standards from your country or region? Please reach out to your Customer Success Manager. We plan to continually update this list based on customer demand.

In addition to local and regional standards, DroneDeploy also supports referencing company-specific standards. This means Safety AI can reference internal guidelines when detecting and reporting risks, offering a more personalized and accurate view of compliance within your organization. For more information on how to add this feature to your projects, please reach out to your Customer Success Manager or support@dronedeploy.com for help.

Manage Risk Filtering

Risk Filtering allows organizations to control which types of safety risks are actively detected and flagged by Safety AI. This customization is ideal for reducing alerts related to risks that are not relevant or prioritized within your projects.

Common types of safety risks that you have the option to filter out are:

A person not wearing gloves
A person not wearing glasses
Fire extinguishers placed on carts
Fire extinguishers not mounted on a pedestal
Cords in the walking path
Material blocking the walking path
Equipment or machinery blocking paths of egress
Equipment, machinery or materials in confined spaces
Missing safety netting along the fall protection system
Material at risk of falling over
Ladder blocking walkway or doorway

By default, all risk types will be detected across every project. To modify this:

Click the dropdown list and select a common safety risk that you would like to filter out.
Click the "Add" button.
Repeat for any additional common safety risks you do not want detected on this project.
Once you have added the risk types you want filtered, click the "Save" button.

Any future walks or panos processed by Safety AI will reference these preferences. This ensures your team is only notified about the risks that matter.

Key features/details

Safety AI customization offers two primary benefits:

Safety Standard Alignment: The ability to cite a specific safety standard allows the reported risks to align directly with your company's daily practices, including local, regional, and company-specific standards.
Prioritized Risk Management: Risk Filtering enables you to suppress alerts for common safety risks that are not relevant or prioritized for a given project, helping your team focus on the most critical issues.

FAQs

Q: Can I use company-specific safety standards with Safety AI?
A: Yes. DroneDeploy supports referencing company-specific standards, which means Safety AI can reference internal guidelines when detecting and reporting risks. For details on how to add this feature, reach out to your Customer Success Manager or support@dronedeploy.com.

Q: What happens if I set a risk filter?
A: When you filter out a common safety risk, it will no longer be actively detected and flagged by Safety AI on that specific project. This ensures your team is only notified about the risks that are relevant.

V2.2