PETRAN: Smarter Asset Performance with AI & IoT

A CMMS records maintenance history and schedules tasks. PETRAN is a real-time intelligence layer above it: it monitors live IoT data, predicts failures with AI, and automatically creates evidenced work orders inside your existing CMMS, making Maximo or SAP proactive instead of reactive.

PETRAN has pre-built bidirectional connectors for IBM Maximo, SAP Plant Maintenance/EAM, Hexagon EAM, Infor EAM, and other major CMMS/EAM platforms. When PETRAN detects a developing fault or inspection defect, it automatically creates a work order in the connected system, attaches sensor evidence and AI diagnostics, and recommends the corrective action. This eliminates manual data re-entry and accelerates maintenance response time from hours to minutes.

PETRAN supports all major industrial IoT protocols: MQTT, OPC UA, Modbus/TCP, BACnet/IP, DNP3, IEC 61850, RTSP/ONVIF, and REST/HTTP. Compatible hardware includes vibration sensors, temperature and pressure transducers, current monitors, flow meters, fixed and mobile IP cameras, thermal cameras, RTLS location tags, and access control systems. Edge agents auto-discover connected devices and normalise data across all sources.

PETRAN's pre-built AI skills for common equipment types (pumps, motors, compressors, conveyor drives, HVAC systems) begin generating predictions from Day 1 using foundation models trained on large industrial datasets. Custom AI models for specific equipment require 30–90 days of historical sensor data to establish behaviour baselines. For assets with limited historical data, PETRAN's anomaly detection capabilities provide value immediately while custom models build accuracy over time.

A standard single-site PETRAN deployment, which includes connecting sensors, configuring edge agents, onboarding users and activating pre-built AI skills for common scenarios, delivers production value within 2–4 weeks of hardware installation. Including discovery, solution design and hardware setup, most teams go from first call to a live pilot in 4–8 weeks. Multi-site enterprise deployments are typically complete within 6–12 weeks. Ombrulla provides implementation services, site surveys, sensor commissioning support, and training for operations and maintenance teams throughout the process.

Yes. PETRAN's edge agents run full AI inference locally on edge hardware (industrial PCs, ruggedised appliances or NVIDIA Jetson devices) without cloud connectivity. They collect, normalise and buffer sensor data locally, so alerts and actions continue during network outages, with sub-second alert latency and local data privacy. Buffered data synchronises with the central cloud hub when connectivity is restored. This makes PETRAN suitable for offshore platforms, remote pipelines, underground assets, and air-gapped networks.

Yes. PETRAN is available in three deployment configurations: SaaS (managed cloud), Private Cloud / On-Premises (customer-controlled infrastructure), and Hybrid (edge AI with cloud orchestration). On-premises deployment supports air-gapped networks, strict data sovereignty requirements, and regulated-industry compliance standards including IEC 62443 and ISO 27001. Containerised deployment via Docker and Kubernetes is supported.

PETRAN implements enterprise-grade security across all deployment modes: AES-256 data encryption at rest and in transit, SSO via SAML 2.0 and OIDC, SCIM user provisioning, multi-factor authentication, role-based access control (RBAC) with granular site-level permissions, and full audit logging of all user and system actions. For on-premises and hybrid deployments, network segmentation and air-gap support are available. The platform aligns with IEC 62443 industrial cybersecurity requirements.

All operational data collected by PETRAN and all AI models trained on your site-specific data remain the property of your organisation. Ombrulla does not claim ownership of customer operational data or derived models. Data is stored in your designated region (UK, EU, US, or customer-specified). Customers can export their data at any time through standard API or bulk export tools. Model portability is supported for enterprise and custom deployment configurations.

Yes. PETRAN's multi-site, multi-tenant architecture is designed for enterprise-wide deployment across multiple plants, depots, construction projects, or infrastructure assets in different regions or countries. Partitioned data architecture ensures each site's data is isolated. Role-based access control allows corporate, regional, and site-level permission structures. Cross-site analytics in the central dashboard enable performance benchmarking and best-practice replication across the organisation.

Industry benchmarks from Gartner, McKinsey, and Deloitte indicate that AI-enabled predictive maintenance reduces unplanned downtime by 30–50%, lowers maintenance costs by 10–25%, and extends equipment service life by 20–40%. Most PETRAN deployments achieve initial ROI within 6–12 months, driven by avoided breakdown costs, reduced overtime labour, and optimised spare parts inventory. Full 5-year NPV typically ranges from 3–5× initial investment.

Getting started with PETRAN begins with a 45-minute discovery call with Ombrulla's solutions team to understand your top asset performance challenges, existing systems, and deployment environment. Ombrulla then provides a tailored demonstration, a solution design proposal, and an ROI model specific to your organisation. Typical time from first call to a live pilot on one asset class is 4–8 weeks. Book a demo using the button at the top of this page or contact Ombrulla at ombrulla.com/contact.