Cement is a low-margin business. The product is heavy, the trucks are expensive, the fuel bills are enormous, and the drivers — paid largely per trip — have every financial incentive to drive fast and brake late. Fleet managers in the industry know this. Most of them have lived with it for years, because they didn't have a way to change it systematically.
A leading cement manufacturer operating across central India came to us in mid-2024 with a specific frustration: they knew their drivers were costing them money and creating safety risk, but they couldn't prove which ones, couldn't quantify it, and couldn't build a coaching programme around gut instinct alone. Their existing GPS vendor gave them location history and the occasional overspeeding flag. What they needed was something their operations team could actually act on.
Their fleet: 350 vehicles, mostly heavy tippers and transit mixers, running between three plants and a network of construction sites across a radius of roughly 400 kilometres.
What the baseline data looked like
Before any coaching or intervention, we spent the first four weeks doing nothing except collecting data. This is a deliberate choice — drivers behave differently when they know they're being measured for the first time, and the novelty effect fades within a month. What you measure in week five is closer to real behaviour than what you measure in week one.
The numbers that came back from that baseline period were not surprising to the fleet managers — but seeing them quantified changed the conversation immediately.
- Average driver score: 54 out of 100. The composite score weighs harsh braking, harsh acceleration, overspeeding, sharp cornering, idle time, and night-driving behaviour. A score in the mid-50s means the fleet is operating well below industry norms.
- 8.3 harsh braking events per 100 km across the fleet average. On a heavy tipper, a hard brake doesn't just wear pads — it destabilises the load, stresses the chassis, and dramatically increases the probability of a rear-end incident.
- 11.2 overspeeding events per driver per week on average, with a long tail: the bottom quartile was averaging nearly 23 per week.
- Idle time at 28% of engine-on hours. On a fleet this size, running diesel engines that burn roughly 3.5 litres per hour at idle, this was a material number before we even looked at driving behaviour.
- 23 minor incidents per month across the fleet — scrapes, kerb contacts, minor collisions while reversing. None fatal, but each one carrying repair costs, insurance implications, and the near-miss shadow of something worse.
The operations head looked at the driver score distribution and said something that stuck: “I could have named you the bottom ten. I couldn't have told you the score was 54 for everyone else.”
The scoring model
ViaLoop's driver score is a composite that runs continuously, not as a monthly snapshot. Each event — a hard brake, an acceleration spike, a speed threshold crossed — is logged with its severity, location, and context. Context matters: braking hard on a highway is different from braking hard in a depot yard. The model accounts for road type, speed zone, and time of day before assigning weight to an event.
Each driver gets a live score visible to fleet managers, and a weekly summary that shows their rank against the fleet, their five worst events of the week (with video-style timeline playback of the sensor data), and a trend line over the past eight weeks.
The fleet manager gets something different: a ranked list, a flag system for drivers whose scores are declining week-over-week, and an aggregate view by route and shift. This last one turned out to be more useful than expected. It became clear early that the night shift, not the individual driver, was the dominant variable in the worst scores.
On the scoring model: We weight harsh braking more heavily than overspeeding for heavy vehicle fleets. A fully loaded tipper doing 62 km/h in a 60 zone is a minor infraction. A fully loaded tipper braking from 60 to 0 in 28 metres is a potential fatality. The weighting reflects the physics, not just the rulebook.
How the coaching loop worked
The operations team ran the first coaching round in month two. They pulled the bottom 31 drivers by score and ran individual sessions — not group lectures, but one-on-one conversations using the driver's own data. A typical session would open with the driver's timeline of their five worst events from the previous week and ask a simple question: “What was happening here?”
Most drivers, when shown their own data, have a reason. Sometimes it's legitimate — a vehicle that cut across them, a pothole they couldn't avoid. Often it's habit: the way they've always braked at a particular junction, the throttle behaviour they developed on an older truck. The conversation becomes about changing a specific habit, not about abstract “safe driving.”
Drivers who improved after coaching were recognised publicly in the monthly fleet meeting — not with bonuses initially, but with acknowledgement. The top five scores each month got their names on a board at the depot. This sounds small. It worked better than expected.
For drivers who didn't improve after two coaching rounds, the operations team moved them to shorter, lower-risk local routes where the scoring consequences of poor behaviour were less acute — and continued monitoring. Six drivers eventually left the fleet voluntarily. The operations head noted that two of those had been problems for years that HR had struggled to address through conventional means; the objective score data changed the documentation picture entirely.
Six months in: the numbers
Driver score: 54 → 76
A 40% improvement in the composite fleet average. More significant than the headline number is the distribution shift: the bottom quartile, previously averaging 38, came up to 61. The coaching disproportionately helped the worst drivers, which is exactly where the safety and cost gains are largest.
Harsh braking: down 43%
From 8.3 events per 100 km to 4.7. This single metric accounts for the majority of the tyre and brake maintenance cost reduction the operator saw — not tracked precisely here, but their workshop lead estimated a 20–25% reduction in brake-related work orders over the same period.
Fuel costs: down 12%
This came from two places roughly equally: idle time reduction (from 28% to 18% of engine-on hours) and smoother driving reducing fuel burn per kilometre. On a 350-vehicle diesel fleet, 12% is a number the CFO noticed without being told to look for it.
Minor incidents: down 39%
From 23 per month to 14. No fatalities in either period, but the reduction in minor incidents matters: each one is a near-miss at some severity level, and the pattern of minor incidents is the leading indicator that serious incidents follow.
Overspeeding events: down 54%
From 11.2 to 5.1 per driver per week. The improvement was fastest in the first eight weeks after coaching began, then levelled out — which is the pattern you want. Behaviour that changes gradually and holds is different from behaviour that spikes down and then reverts.
What surprised us
Experience didn't predict score. We expected the youngest, least experienced drivers to be at the bottom of the distribution. They weren't. Three of the ten worst-scoring drivers had over twelve years with the company. Long tenure had entrenched habits that newer drivers, more recently trained, didn't have. The coaching effort actually went hardest against the most senior people — which required careful management of the conversation.
The night shift was a route problem, not a driver problem. When we broke scores down by shift, the night shift came in consistently 11–14 points lower than day shift across the same drivers. The reason, once we looked at the route data, was straightforward: the night routes included a stretch of poorly lit state highway where every driver braked harder and more erratically due to visibility. The operations team rerouted night vehicles around that stretch. Scores on the night shift closed the gap within three weeks — without coaching a single driver.
The score board created peer pressure the operations team couldn't manufacture themselves. Within two months, drivers were asking to see their score before their manager asked to discuss it. A handful of the higher-scoring drivers started coaching newer colleagues informally. That dynamic — peer-to-peer behaviour transfer — is difficult to engineer top-down and it emerged on its own.
Where they are now
The operator has rolled ViaLoop Fleet to a second plant in a different state, adding 110 vehicles to the programme. They're currently working with us on a driver-facing mobile app that gives each driver their own score, trend line, and weekly event summary — shifting some of the feedback loop from the fleet manager's desk directly to the driver's phone.
The operations head put it plainly in a recent review: the ROI conversation stopped being about fuel savings after month three. The number that moved the most senior stakeholders was the minor incident rate. In an industry where a serious accident can shut down a plant, affect insurance premiums for years, and carry personal liability for the operations team, a 39% reduction in the leading indicator matters more than the fuel bill.