Problem

Mancini’s Bakery is a family-owned Pittsburgh institution with decades of tradition. While the bakery has invested in new production lines more recently, prior to this project, Mancini’s managed production entirely by hand:

• Batch records logged on paper
• Ingredient lots tracked by memory and manual logs
• Virtually no system connecting raw ingredients to finished products

A supplier recall under the old system would have triggered hours of manual investigation with no guarantee of completeness. This not only made inefficiencies undetectable but also created a hard ceiling on growth.

BRCGS (Brand Reputation through Compliance Global Standards) certification is a prerequisite for supplying major retail chains and foodservice distributors. It requires documented, auditable, forward-and-backward lot traceability across the entire production process. Without certification, Mancini’s could not compete for contracts that would allow the bakery to grow.

Digitizing a largely hand-operated bakery presented MDI with challenges that went well beyond a typical SCADA or MES deployment. Mancini’s had no conveyance or automated methods for distributing materials around the facility. Human beings transported everything from raw ingredients, troughs of dough, oven racks, and packaged products from place to place.

This led Mancini’s to invest in RFID technology. The idea was that all vessels that move dough through the facility would utilize automatic RFID tracking for traceability. However, RFID technology has its own set of challenges.

RFID readers deliver data through arrays of complex JSON objects that are spread out across multiple OPC tags. Every single tag scan sends a large, multi-layered payload that requires ASCII conversions as well as additional context in order to interpret each scan. Each array instance represents a single RFID tag picked up by an antenna. Scans are concentrated at the top of the array, which means the position within the array of the asset being scanned is inconsistent. The readers will also occasionally lose all RFID tags for a few seconds before the scan data comes back online. This created a number of edge cases and challenges.

MDI also needed to account for manual tag configuration. A system that required operators or engineers to manually create and map tags for every new RFID device scanned into the system could not scale properly and would introduce the same human-error risk Mancini’s was trying to eliminate.

MDI had to integrate and translate this data into a human-readable format and deliver the data seamlessly to Kanoa MES’s Trace Module in order to provide full genealogy graphs without any gaps in the data. This required a robust, but lightweight and dynamic software translation layer built on Ignition.

Solution

The solution’s defining innovation is how MDI leveraged Ignition’s Event Streams Module. This module serves as the architectural backbone of the entire traceability system.

Taming RFID Data with Event Streams

RFID readers positioned at critical control points throughout the bakery — receiving, dough mixing, proofing, baking, and packaging — stream read events into Ignition as arrays of JSON dictionaries. The Event Streams Module processes these raw payloads in real time, parsing each JSON array, extracting individual tag reads, and transforming them into clean, structured event records. This solved the core data quality problem: what arrived as noisy, unstructured RFID output left the pipeline as organized, contextualized production events.

Automatic UDT Generation

MDI used Event Streams to automatically generate UDT instances whenever the system detected a new RFID device. When the system recognizes a previously unseen tag, the pipeline creates the corresponding UDT instance on the fly, defines the device as a Kanoa asset under the correct group, and then populates scan data without any manual engineer intervention. This eliminated a major source of manual configuration overhead, ensuring the tag structure remained consistent and scalable as the system grew.

Integration with Kanoa MES Trace

Enriched, structured RFID events flow from Event Streams directly into the Kanoa MES Trace Module. Kanoa uses these events to build complete traceability graphs for every production batch. This data includes which raw material lots went into each batch, how dough moved through each stage, and which finished products were created as a result. The graphs support both forward traceability (given a raw material lot, which finished products contain this raw material) and backward traceability (given a finished product, which raw material lots contributed). This provides the audit documentation required for BRCGS certification.

Operators interact with the system through Perspective-based dashboards that display live lot movement, active device status, and traceability graph summaries. MDI designed the interface for bakery operators, not engineers, limiting complexity while the Event Streams pipeline handles the heavy lifting “behind the scenes.”

MDI engineered the system to be fully capable of correcting errors, allowing humans to make changes, and providing the MES with context on out-of-process events. The system also needed to account for raw ingredient receiving and movement since RFID tags were too costly. MDI created multiple different interfaces to allow users to interact with the system while still minimizing clicks.

Other Digital Transformation Efforts

Through this effort, MDI also combined their internal UDT framework for quick-to-deploy dashboards with Kanoa’s dashboarding and marquee tools. Both MDI and Mancini’s can quickly link any data point that represents either a metric or a state to intuitive widgets without needing to write custom code or spend days developing complex visuals.

During the first phase of implementation, MDI designed and deployed a dashboard that helped bakers adjust machine settings within one hour. This dashboard brought together data points spread across multiple HMIs and stack lights into a single, easy-to-understand marquee. Something as simple as seeing the real-time level of the hopper has already improved operational efficiency since the bakers no longer have to wait for the “hopper empty” alarm to know when it needs to be filled.

Result

The impact of this implementation has been direct and transformational. A family-owned business that previously could not have competed for major retail contracts is now on a clear path to BRCGS certification. The system provides full forward-and-backward lot traceability in real time and in audit-ready format through Kanoa’s traceability graphs.

Beyond certification, the project delivered something Mancini’s was always missing — visibility. Management can now see how materials move through the facility in real time, understand the full scope of what ingredients they have on hand, and respond to supplier quality events with speed and confidence that was simply not possible under a paper-based system. What once would have taken hours of manual investigation now takes seconds.

Perhaps most meaningfully, this project proves that the Ignition Effect extends to small, traditional businesses, not just large industrial operations. The flexibility of Ignition’s platform and the power of the Event Streams Module made it possible to bring enterprise-grade food safety technology to a family bakery, on a timeline and at a cost that would have been impossible with traditional MES or ERP approaches.

Project Team

This project started out as a de-risking effort by the Digital Foundry at New Kensington before being handed off to MDI. The Digital Foundry’s team consisted of three engineers (one senior level, two junior level). MDI’s team consisted of Greg Schneider, a Gold Certified Ignition veteran, and Sanskar Lamsal, an aspiring junior engineer who brought a year of intense, MES-controlled run schedule experience to the team. MDI also worked closely with Brady Cummings at Mancini’s Bakery to ensure all device connectivity, networking and servers were stood up properly. Mancini’s Bakery was not afraid to get their hands dirty and worked closely with MDI every step of the way to ensure success and a clean transition.

Project Scope

• Start Date: January 2026

• Deploy Date: Phase 1 - February 2026; Phase 2 - April 2026; Phase 3 - July/August 2026

• Tags: 15,600+

• Screens: 50+

• Clients: 50+

• Alarms: 100+

• Devices: 15 Turck RFID readers; 12 other PLCs and devices such as Rockwell PLCs, Siemens PLCs, Schneider Electric PLCs, BACnet Devices, etc.

• Architecture: Standard, single Gateway

• Databases: 1 MySQL and 1 MSSQL

• Historical data logged: 11.4M events processed per month; 4.4M tag history total records to date

• Number of people on team: 2-3

• Ignition Ecosystem: Kanoa MES