Corporate AI8 min read

How Can Fleet Managers Use AI?

On this page

Fleet managers can use AI to turn trip and fuel data into clear reports in Excel, build maintenance schedules from the manufacturer's service intervals, write short driver messages and summarise incidents for management or the insurer. AI is fast at sorting numbers, spotting odd patterns and writing plainly. It does not know your vehicles, so the figures come from your tracking system, fuel records and workshop, and a person checks every conclusion before anyone acts on it.

This guide is for fleet managers, transport officers and operations staff in Kenya who run anything from five vans to a hundred trucks. For the wider picture of AI in freight, clearing and the Northern Corridor, read AI for logistics and transport companies in Kenya. For matatu SACCOs and small operators, see how a transport business or matatu SACCO can use AI.

Where a fleet manager's time goes

Much of a fleet manager's week is paperwork: exporting trips from the tracking portal, matching fuel receipts, chasing service dates, answering "where is my truck?" and writing up what went wrong. Most of it is copying, sorting and summarising. AI does that well, as long as the numbers going in are right.

Six tasks where AI helps, with prompts

Each prompt uses the four Cs we teach at FuKazee: Clarity, Context, Constraints and Customisation. Before you paste any export, replace driver names with codes (D01, D02) and plates with fleet numbers (V01, V02).

1. A weekly trip report from your tracking export

I manage a fleet of [number] vehicles in Kenya. Below is a weekly trip export from our tracking system with drivers and plates replaced by codes. Columns: vehicle code, driver code, date, start time, end time, distance in km, idle time in minutes. In Excel, show me how to build a summary table per vehicle with total km, number of trips, average trip length and total idle time. Then list the five vehicles with the highest idle time as a share of driving time. Give the formulas, not just the answers.

Check: total km in your summary matches the tracking portal's own total. If it does not, find the missing rows before you trust anything else.

2. Fuel use by vehicle and route

Using the fuel log and trip data below (codes only), calculate km per litre for each vehicle for the month. Flag any vehicle whose km per litre is more than [15]% below its own three-month average, and any fill-up larger than the tank size of [number] litres. Present a table and three sentences on what to look into first. Do not suggest causes I have not given you.

Check: a low figure is a question, not proof of fuel theft. Check the receipts, odometer readings and the route (hills, traffic, loads) before you speak to anyone. Our Excel AI training page covers this kind of analysis.

3. A maintenance schedule from the service intervals

Below are the service intervals from the manufacturer's manuals for our three vehicle models, and each vehicle's current odometer reading and last service date. Build an Excel schedule showing the next service due for each vehicle by km and by date, whichever comes first. Add a column that turns red when a service is due within [1,000] km or 14 days. Use only the intervals I give you.

Check: the intervals must come from the manual or your workshop, never from AI's general knowledge. Your mechanic signs off what each service includes.

4. Driver messages that are safe to read

Write a short message to our drivers, in simple English and in Kiswahili, about [the new rule that all vehicles must be back at the depot by 8 pm from Monday 7 December]. Under 50 words each. Start with the instruction, then the reason in one line. Add a final line: "Read this when parked. Never use your phone while driving."

Check: a supervisor who speaks Kiswahili reads it. Send messages at times when drivers are likely to be stopped, and never ask for a reply while a vehicle is moving.

5. An incident summary for management or the insurer

Below are the confirmed facts about an incident involving vehicle V07, with names removed: the driver's statement, the tracking data for the hour before, and the workshop's damage note. Write a one-page summary for the general manager with: what happened (facts only), timeline, damage and cost estimate as given, what we have done, and open questions. Keep each source's facts separate. Do not decide who was at fault.

Check: the timeline matches the tracking data. Fault, claims and any discipline are for the people responsible, with the insurer and police where needed.

6. A monthly fleet report the board will read

Using the monthly figures below, write a one-page fleet report for the board. Include: fleet availability, km driven, fuel cost per km, maintenance cost, incidents, and the three decisions we need from management. Plain English, short paragraphs, one small table. FIGURES: [paste]

Check: every number traces back to a sheet you can show. If the board asks "where did this come from?", you should be able to answer in one click.

What AI must not do for a fleet

  • Judge drivers. AI can sort data. Deciding that a driver is careless, dishonest or should be disciplined needs facts, a fair hearing and HR, not a chatbot's summary.
  • Diagnose mechanical faults or clear a vehicle as safe. That is your mechanic's job.
  • Hold personal data. Driver names, ID numbers, licence numbers, phone numbers and exact home locations stay out of public AI tools.
  • Encourage phone use on the road. Messages, quizzes and "quick replies" must never be designed for a driver who is moving.
  • Tell you the law. For licensing, inspection and road rules, go to NTSA and your legal adviser.

Tracking data and the law in Kenya

Location data can identify a person. The Data Protection Act, 2019 defines an "identifiable natural person" as one who can be identified directly or indirectly by reference to an identifier such as a name, an identification number, "location data" or an online identifier. A trip log tied to one driver is that driver's personal data.

The Office of the Data Protection Commissioner (ODPC) published Guidance Notes for the Transport Sector in 2026 (the foreword is dated 30 June 2026). The note says surveillance tools, "including CCTV, dashboard cameras, and vehicle tracking systems", can improve safety but may lead to excessive monitoring without clear policies on data access, retention and use. It also tells transport operators to list outside providers who handle personal data, such as GPS tracking vendors, and to confirm that data processing agreements are in place. Its worked example records fleet tracking data with legitimate interests as the lawful basis and a set retention period.

The Data Protection (Registration of Data Controllers and Data Processors) Regulations, 2021 also list "transport services firms" among those that must register with ODPC. Check your company's status with your adviser.

What this means for AI: export the data you need, mask it, analyse it and delete the copy. Do not upload a full driver database to a public tool. Our guide can staff use ChatGPT under Kenya's Data Protection Act? explains the rules in plain words.

This is general information, not legal advice.

How FuKazee trains transport teams

FuKazee trains dispatch, accounts, safety and customer staff on their own kind of paperwork. Our logistics training page says we trained two teams on-site at A.O. Bayusuf and Sons, a Mombasa transport company, in October 2026. It also says sessions can start early and run in split groups, so dispatch and the yard stay staffed. We show the free version of each feature first, plus Copilot for teams on Microsoft 365.

Questions people ask

Can AI connect directly to our tracking system?

Some fleet platforms offer their own AI features. A general chatbot does not plug into your tracking system by itself. The simplest safe start is to export a report, mask it and work on the copy.

Which AI tool is best for fleet reports?

The one your team already has and is allowed to use. If you work in Excel on Microsoft 365, Copilot may be available inside it, depending on your plan. Free chat tools can also explain formulas you then build yourself. Read how to use Copilot in Excel for a step-by-step start.

Can AI predict when a vehicle will break down?

Not reliably from a chatbot. It can help you spot vehicles with rising repair costs or overdue services from your records. Real prediction needs good data over time and a proper system, and your mechanic still decides.

Is it legal to use dashcams that face the driver?

ODPC's transport guidance warns that monitoring tools can become excessive without clear policies. Before installing driver-facing cameras, agree the purpose, retention period and who can see the footage, and get legal advice.

Will AI replace fleet managers?

No. AI takes the copying and summarising. Deciding which vehicles to retire, which routes to change and how to treat drivers fairly stays with a person who knows the fleet.

Sources

All sources last checked 15 November 2026.

Where FuKazee fits in

FuKazee is a Nairobi training company that has trained more than 3,000 people since 2022. We help fleet and transport teams turn their own trip, fuel and workshop records into clear reports with AI, safely. Tell us how many people would attend and which reports take up most of their week. Email lead@fukazee.com

Want guided support instead?

Plan AI training around your team's real work. Answer a few quick questions about your sector, team size and goals, and we will recommend the right programme and format. Sessions are built on your workflows and include a usage policy.

Rodney MainaFounder & Director, FuKazee

Rodney has run FuKazee's training since 2022 and still teaches inside every programme line: corporate sessions, the postgraduate cohort and the KCSE examiner webinars. First Class Honours in Journalism and Mass Communication, University of Nairobi.