Client: An organization dedicated to helping builders of intelligent devices, data analytics, applications and services to enable delivery of healthcare services more efficiently and with better outcomes.
Situation: Seeking a more cost-effective method for semi-automated inspection of parenteral drugs that doesn’t require human input and decision making
Misfire: Client attempted to retrofit a semi-automated system that leveraged traditional computer vision and deep learning resulting in an inaccurate 30% rejection rate.
Solution: Boon Logic retrofitted the organization’s semi-automated inspection equipment with AVIS, the only AI-based visual inspection solution that self-trains models from “normal, defect free product” to detect anomalies.
- 88% fewer false rejections than with humans improves inspection accuracy.
- 22% increase in output without intermittent stoppage improves production capacity.
- Reduces need for 2.5 persons per retrofitted machine.
Client: Great Lakes Dredge and Dock is the largest provider of marine dredging in North America.
Situation: The company seeks technical innovations that allow its barges to efficiently protect the nation’s shorelines and mitigate potential risks associated with storms and sea change. By automating anomaly detection with analytics that provide greater clarity about equipment deviations, the company seeks to gain early indications of equipment failure to ensure crew productivity.
Solution: Boon Logic’s Amber AI-based predictive maintenance and condition-monitoring solution reduces the likelihood of unexpected operational failure and provides crew members a root cause of the failure taking place for quick and precise action.
- Completed pilot within 90 days, far outpacing traditional machine learning solutions that require more lengthy modeling
- Custom solution for commercial deployment developed and delivered within 3 months.
- Continuing rollout to its 33-barge fleet.