We’ll walk you through the platform and explain the capabilities and ways it could benefit your business. The goal is for us to learn about your needs and for you to learn about how HoneyBrain can solve them to determine if its a good fit to move on to the discovery phase.
The HoneyBrain platform comes in two modes, Workstation and Enterprise. We recommend that all our clients start by using Workstation, then move on to implementing Enterprise. You can read more about Enterprise mode below.
Workstation mode runs all of its inspections on a standalone workstation per use case. Typically, a Workstation implementation is around $75k. This is a big advantage HoneyBrain has over completely custom solutions as you can see a proof-of-concept relatively quickly and easily to ensure that a larger investment is worthwhile. This stage also makes the Enterprise implementation phase smoother.
Once HoneyBrain has been successfully implemented in Workstation mode, the platform can be upgraded to Enterprise mode and more use cases and data integrations can be added.
Since HoneyBrain is a platform, the functionality is configured on a per use case basis. Configuration defines information such as the use case, the inspections to run, and how to communicate with the outside world for triggers, cameras, BOM/Order Data, and quality systems.
The development effort that is unique per use case is the AI pipeline and AI model, but specific steps are taken to try to get as much re-use as possible. We recognize that many customers have multiple machines performing the same process and sometimes this goes across plants and even continents. We pre-process the media and train our models with the goal of reusing most of the effort across stations.
The platform can be run in Training, Test, or Production mode. In Training mode, we are capturing media to use in the AI model. In Test mode we start running inspections on that media to verify the AI model is working properly, but it does not write data to your quality system. That only happens when we turn on Production mode and the platform is fully operational.
The HoneyBrain platform comes in two modes, Workstation and Enterprise.
Workstation mode runs all of its inspection on a standalone workstation per use case. Triggers, BOM/Order Data, and quality systems would indirectly interface with the workstation.
Enterprise mode integrates into the customer’s information systems more closely and more directly with the intention of providing a platform across multiple use cases throughout one or more customer facilities. This mode can be used to run some use cases without a line-side/edge computer. A camera can be installed, and the server captures media directly from this camera. Components can be hosted on premise or in the cloud depending upon the customer’s needs. Typically, in Enterprise mode, lighter weight line-side/edge computer installations make upgrading the software or bug fixes easier because of the more centralized approach.
Remote configuration — Enterprise uses a centralized environment to maintain all configurations versus workstation mode where changes to the inspection configuration are made directly on the workstation.
Integrated authentication – Access and roles can be integrated with the customer’s authentication provider.
Multiple environments (such as dev, staging, and production) – allow the ability to run inspections in parallel with production. The trigger(s) and media are captured by the production system, but a draft AI model or configuration change can be running parallel to the enterprise environment.
Media and inspection search — provides a repository that stores the inspection and media with a user interface able to search and find previous inspections. The search can also be extended to other vision systems (even systems not part of the platform) within the plant allowing the other systems to push the inspections and media into a consolidated search and reporting environment.
Monitoring and analysis — provides performance metrics, monitoring, consolidated log data, and other operational support functionality.
Notifications — proactive messages from the platform indicating something of interest. For example, an error reaching a camera. These notifications are written to a database with a search user interface but can also be extended to send to ticketing system(s), email, or other systems made available by the customer.
