B2B SAAS Dashboard (Part II)
After fixing live trip visibility and geofencing, we turned our focus to what happens during trips: how drivers actually use vehicles, & how to build systems that improve vehicle uptime without adding friction.
TL;DR
Company: Alt Mobility (EV leasing & asset management)
Product: MyFleet — a dashboard for fleet operators
Users: City managers, fleet founders, ground staff managing 200–700 EVs
Problem: Poor visibility, Lack of control, inefficient method of monitoring drivers
Solution: Real-time vehicle tracking, trip & stop categorisation, geofencing alerts
Impact: Our customers saved an estimated ₹1.3–1.6 crore ($200k) monthly
What does Alt do?
We help fleets lease commercial electric vehicles (EVs) across 20+ Indian cities. With ₹200 cr worth of vehicles to manage, real-time monitoring is non-negotiable to keep them running smoothly.
Seen Zepto, Zomato EV drivers riding flashy two-wheelers? Or that three-wheeler cargo pickup waiting outside your house with a delivery package? Remember the last time you booked a vehicle via Porter to pick up your stuff? Yes, I’m talking about those vehicles.
Problem Statement
Managing a fleet was a manual and inefficient process. Operators struggled to ensure driver accountability, prevent misuse, and respond to emergencies like theft until it was too late.
No insight into driver behavior: There was no way to monitor how drivers were using vehicles. Rash acceleration, overspeeding, and traffic violations often went unnoticed, until penalties (challans) or breakdowns occurred.
Low vehicle uptime due to misuse: Harsh driving habits led to more wear and tear, reducing the lifespan of EVs and increasing downtime.
Delayed theft response: In the rare event of vehicle theft, operators relied on manual recovery teams, leading to delayed action and higher risk of vehicle loss.
These inefficiencies eroded trust between fleet owners, supervisors, and drivers—making it harder to maintain safety, efficiency, and control at scale.
Objective
Our goal was to design tools that could help fleet operators:
Improve vehicle uptime by spotting vehicle issues early.
Give better visibility into driver behaviour, reward good driving, and act fast when things went wrong.
My Role
👋🏼 Hi again! This was my second phase of MyFleet project, where I led design for a set of safety & performance tools:
Design and test custom alerts & driver scoring models
Prototype early immobilisation flows for vehicle theft or risk
Research Findings
In over interviews with fleet owners, city managers, and ground teams, we uncovered recurring pain points:
Operators had no structured way to measure or improve driving behaviour.
There was a manual & unclear feedback loop between driver activity and fleet decisions (like incentives or warnings).
Since light commercial EVs were light weight, vehicle theft had become common in certain areas.
Most actions were reactive, not proactive. Operators would spend more time fire-fighting.
Symptoms got fixed. The root cause didn’t. (Operators focused on breakdowns, not behaviour)
We noticed a pattern during our shadowing sessions: the top-performing drivers by delivery count (incentive), was also the one most likely to push the vehicle beyond safe limits. They'd over-speed, brake harshly, and accelerate aggressively to maximise output. The damage wouldn’t show up immediately but six weeks later, when the vehicle broke down and needed servicing.
Our takeaway was clear: without visibility into behaviour, performance metrics alone were misleading.
— Fleet Operator, Delhi-NCR
User Stories
We condensed the above problems and insights from our research into simple user stories:
As a fleet operator, I want to set custom alerts (for over-speeding, idle time, harsh braking, etc.) so that I can proactively monitor risky driver behaviour and prevent accidents.
As an operations lead, I want to see a daily driver score based on driving patterns so that I can identify top performers.
As a fleet operator, I want to remotely immobilise a vehicle in case of theft or misuse so that I can prevent further loss or damage.
Solution (Design) — Introducing MyFleet!
We first needed to understand the data from the IoT device and its assess its accuracy. While we received a lot of data, not all of it was useful. Our first goal was to identify the most relevant data points for our customers and design the interface to highlight these points visually.We worked with the data science team to assess the accuracy of the data and prioritise the most important parameters. We used a radar graph (below) to cherry pick the most relevant data points.
We began with paper wireframes and iterated multiple times before making wire-frames on Figma. This process helped us pinpoint exactly what the operator would need and how to provide that information with the least amount of friction.
Introducing Customisable Alerts Tags to assess Driver Behaviour
PROBLEM
#1
No insight into driver behavior: There was no way to monitor how drivers were using vehicles. Rash acceleration, overspeeding, and traffic violations often went unnoticed, until penalties (challans) or breakdowns occurred.
After observing the daily routine of our user, the team agreed that out of all the possible alerts, only 4 were crucial for daily monitoring. We allowed the operator to set thresholds for:
Over-speeding
Harsh Acceleartion
Harsh Braking
Excessive distance covered (typically >180 kms)
We realised each vehicle category (for eg. 2W vs 4W) and operator had unique needs. Instead of using universal alerts, we allowed the users to set custom thresholds.
Driver Score
PROBLEM
#2
Low vehicle uptime due to misuse: Harsh driving habits led to more wear and tear, reducing the lifespan of EVs and increasing downtime.
We wanted to make use of the vast amount of data we were receiving every 30 seconds to give insights. To help operators evaluate driver performance, we analysed vehicle alerts (e.g., over-speeding), challans, excessive distance covered, and geo breaches to assign the driver/vehicle a score from 1 to 10 with colour-coded severity levels.
0–4 = Poor (Red)
5–7 = Satisfactory (Yellow)
7+ = Good (Green)
This scoring system would help the fleet operator make quick decisions in handing out weekly cash incentives. How we calculated this score would be an ongoing project to improve.
Giving operators the control of their Vehicle by Remote Immobilisation
PROBLEM
#3
Delayed theft response: In the rare event of vehicle theft, operators relied on manual recovery teams, leading to delayed action and higher risk of vehicle loss.
In our early testing, we discovered that fleet tracking platforms need a secure way to control vehicle operations remotely. The idea behind immobilisation is to let users disable or enable a vehicle from running with a confirmation prompt.
Every screen in this flow was critical and needed to be carefully designed, as any mistake could lead to life-threatening consequences if not handled with proper precaution. We revisited multiple flows and exhausted all possibilities before adding an OTP-based verification to prevent authorised usage.
We tested this flow on one of our vehicles for six weeks before launching it.
Usability Testing
We didn’t run formal usability tests, but we validated critical flows through internal feedback and real-world testing:
We tested the immobilisation flow in real-time on a live vehicle over a 6-week period. This helped us identify risks, edge cases, and moments of hesitation that wouldn't show up in static screens. Based on this, we introduced an OTP confirmation step and warning prompts to prevent misuse.
Feedback from PMs, engineers, and the ops team helped us simplify alert customisation and clarify how the driver score was being calculated.
Shadowing sessions gave us insight into how operators scanned the dashboard under pressure—leading to layout and label tweaks for faster readability.
Google Analytics & Hotjar helped catch UI friction points in early rollouts.
Even without a formal test plan, these hands-on feedback loops helped us design something that felt safe, intuitive, and actionable for operators on the field.
Impact on Business & KPIs
By solving core issues around driver behavior and vehicle misuse, we helped fleet operators gain more control and confidence in their operations. Manual firefighting was replaced with data-backed decisions, and safety became a shared responsibility across the fleet. The results were clear:
Operational efficiency: Reduced collection TAT by ~40%, thanks to real-time location access and remote immobilisation.
Improved safety: Immobilisation was used 300+ times in a month, preventing multiple vehicle thefts, saving about 90lakh - 1.2cr per month.
Adoption & engagement: 75% of fleet operators now use MyFleet daily, making it a core part of their workflow.
Revenue impact: As operations became smoother, customers scaled their fleets—leasing more vehicles from us and directly contributing to our growth.
Their success became ours, proving that when we solve the right problems, everyone wins.
Team :)
This second phase of our discovery started with a simple goal: understand how drivers actually use their vehicles, and how we could design better tools to support them. What followed was weeks of listening, shadowing, and digging into the messy details of EV fleet operations. I wasn’t just designing screens, I was learning to solve the right problems.
Again, I couldn’t have done this without the help of my collaborative high-agency team. A big thank you to the incredibly talented people for helping me in every step of the way.
Rohit Kashyap (UX Design & Research Intern)
Yashoheet Sethi (Product Manager)
Prateek Jha (Full-stack developer)
Sharad & the entire Data Science team.
+founders, engineers, operations team. And to our customers who took out time to sit with us and graciously offered us to shadow them :)
