Twenty-nine years between the lakes and the border

Mandate

Create the company's first data function, with authority over data shared with shippers, customs authorities and customers' emissions reporting, and over the forecasting used to plan the fleet and terminals.

Outcomes

• Moved arrival communication to windows with a published calibration record; 88% of 1,510 arrivals fell inside the window in 2025.
• Waiting-time claims paid fell from C$2.9m in 2021 to C$1.1m in 2025.
• Introduced a season-close rule; no unplanned icebreaker assistance at the 2023, 2024 or 2025 closures.
• Placed 34 shippers on one data-sharing agreement built around permitted questions.
• Issued fuel-based, ballast-inclusive emissions statements to 41 marine customers from the 2025 season.

Decisions in context

Rémi Vautrinel created the chief data officer role after the 2019 ownership change, when our new investor asked questions about fleet performance that took weeks to answer. My mandate was wider than reporting. I own the data platform, data governance and privacy operations, forecasting for the fleet and terminals, the shipper portal and the data behind customs filings and emissions statements. I do not own infrastructure or cyber security, which sit with our chief information officer, or the company's own emissions inventory, which belongs to our sustainability lead. The work has been to make the data that leaves the company dependable: arrival windows with a published record, a season-close rule, one agreement for shipper access and measured emissions figures. I present twice a year to the board's audit and risk committee on data governance, privacy and forecast performance, and I am not a director.

Reason for the move

Current appointment; completing the platform programme and road-division emissions statements by mid-2027.

Mandate

Bring the fleet's voyage, fuel, lock and terminal data into one platform and give operations and commercial teams figures they could agree on.

Outcomes

• Replaced daily noon-report fuel estimates with continuous flow-meter readings on all fifteen vessels by 2019.
• Reconstructed eleven seasons of lock transit times from vessel logs and positions, later the basis of arrival windows.
• Cut ballast distance by about 6% through a voyage-pairing tool used by the commercial desk.
• Prepared fleet performance data for the 2019 investor transaction's data room within five weeks.

Decisions in context

I came back to Morvaise to solve a problem I had left behind in 2006: nobody could say with confidence how much fuel a voyage had burnt or how long a lock passage had taken, because the figures lived in noon reports, logbooks and terminal spreadsheets kept in three formats. With our technical superintendents I installed and calibrated fuel flow meters, linked them with engine data and transponder positions, and reconciled the results against bunker receipts until the chief engineers trusted them. I also reconstructed lock transit times for eleven seasons. The commercial team used the platform first for voyage pairing, matching downbound and upbound cargoes to shorten ballast legs. In 2019, when the family sold a minority stake, the data room for fleet performance was built from that platform. The investor's diligence questions persuaded the chief executive that the company needed a data officer.

Reason for the move

Appointed the company's first chief data officer in February 2021.

Mandate

Turn the carrier's dispatch, fuel and border data into decisions on pricing and network design, and lead the data side of the camera and telematics programme.

Outcomes

• Wrote the purpose register and retention limits for camera and telematics data across 1,240 tractors in 2015.
• Preventable collisions per million kilometres fell from 0.84 to 0.57 in the two years after installation.
• Reduced empty kilometres from 17% to 13% of the total through lane pricing and backhaul matching.
• Built the lane pricing model used for every contract renewal from 2013.

Decisions in context

This role took me from planning loads to deciding what the company should do with the data its trucks produced. In the first two years I built lane pricing and backhaul matching, which reduced empty running and gave the sales team a defensible basis for renewals. From 2014 the camera and telematics programme dominated my time. I chaired a working group of safety, operations, human resources and two elected driver representatives that decided what data we would collect, who could see it and when it would be deleted, and I negotiated the insurer's access terms with our broker. The register we adopted in 2015 was later applied to fuel cards, dock appointment systems and the driver mobile application. When the Thibervaux family sold a majority stake to a US carrier in December 2016, the new owner planned to run analytics from its own headquarters. I helped transfer the work and left in February 2017.

Reason for the move

After the 2016 sale, analytics moved to the new owner's US headquarters, and Morvaise invited me back to lead its fleet data.

Mandate

Reduce border delays caused by manifest errors and rebuild network planning for the cross-border lanes.

Outcomes

• Moved shipment data capture to order intake; document-related delays fell from 6.8% to 1.4% of southbound crossings by 2010.
• Raised on-time delivery on cross-border lanes from 84% to 92%.
• Built load planning that accounted for drivers' remaining hours under both Canadian and US rules.
• Brought 112 shippers onto electronic order data by 2011, up from 9.

Decisions in context

Thibervaux hired me because its border problem looked like a technology problem, and it turned out to be a problem of where data originated. In my first months I followed shipments from order to crossing and found that most manifest errors were created at the dispatch desk, from paper documents, minutes before departure. I proposed capturing data at order intake and refusing cross-border dispatch without it, and after a trial on two lanes the founder adopted the rule. The six border data coordinators I inherited stopped retyping documents and began helping shippers set up electronic orders, in French and English. I also rebuilt load planning so that drivers' remaining hours under Canadian and US rules were part of the plan rather than a dispatcher's memory. The work cost us one large customer and gave the company a reputation with customs brokers for clean data.

Reason for the move

Promoted to vice-president, operations analytics.

Mandate

Plan voyages for the fleet and, from 2004, run the scheduling team.

Outcomes

• Started the company's first record of lock transit times from vessels' own logs, kept every season from 2002.
• Introduced a mid-November review that fixed each vessel's final voyage before closure.
• Reduced berth clashes at the Hamilton terminal by sequencing arrivals against loading rates.
• Completed a part-time master's degree in 2005 with a thesis on queueing at the canal locks.

Decisions in context

Morvaise's fleet operations director had read my ice advisories for years and hired me to plan voyages. I worked between the commercial desk, which fixed cargoes, and the vessels, which knew what was actually possible. Their views of arrival time often differed, and the commercial desk's usually reached the customer. I began recording lock transit times from our vessels' logs so that at least one input to the schedule was measured rather than remembered. From 2004 I managed the five schedulers and introduced a review in mid-November that fixed each vessel's final voyage before closure, an early version of a rule I would return to almost twenty years later. My master's thesis, on queueing at the canal locks, grew out of that work. By 2006 I wanted to understand freight on land and at the border, and Thibervaux offered me the opportunity.

Reason for the move

Wanted experience of land-side freight and border operations; recruited by Thibervaux Transport.

Mandate

Prepare ice and weather routing advisories and passage-time estimates for client vessels.

Outcomes

• Kept a verification log of every passage-time estimate against the actual passage for three and a half winters.
• Showed that the firm's estimates were well centred but too narrow, and introduced stated ranges in advisories from 2000.
• Prepared winter advisories for Morvaise Navigation among other lake and ocean clients.

Decisions in context

This was my first job after university, and it shaped most of what I believe about forecasts. Brumelle advised ship operators on routes through winter ice and storms: which passage was open, how long a crossing would take, when to wait. I prepared advisories from ice charts, satellite imagery and weather models, and estimated passage times for masters who would plan their next days around them. After a badly underestimated passage in my first winter, my chief forecaster asked me to log every estimate against the result. The log showed that our estimates were centred well enough but stated with far too much confidence. From 2000 our advisories included a range, which masters told us they preferred. I left when Morvaise, which had received our advisories for years, asked me to plan its voyages.

Reason for the move

Recruited by Morvaise Navigation, a client, to plan its voyages.