Exception Management
Data lakes & Machine Learning based KPIs management
The exception management module manages exceptions and prevents process(es) from breaking down by using different components of the monitoring pipeline. Machine Learning (ML) technology is proactively trained on users' behavior through numerous sources that can be plug-and-play in nature. It continuously checks the incoming stream of data for any abnormal activity.
Search exceptions, trips, KPIs⌘KException monitor
Dubai · 28 May · 5 sources streaming
Pickup dwell · T2
45 min
Transit JEA → KEZAD
1h 38m
Idle trucks
3
Open exceptions
1
| Exception | KPI | Observed | Status |
|---|---|---|---|
| EXC-000236TRIP000201 · 09:48 | Idle time | 52 min · Mussafah | Escalated |
| EXC-000233TRIP000198 · 08:21 | Route deviation | +14 km · E311 | Resolved |
| EXC-000231TRIP000195 · 07:12 | Reefer temperature | 7.8 °C · KEZAD | Resolved |
Key features
What Exception Management gives your team.
- Telematics
- Terminal gate log
- Driver app
- ERP orders
Telematics10:58:12
TRK-08 stationary · Jebel Ali T2
Gate log10:58:40
TRIP000215 gate-in · no gate-out
Driver app10:59:05
Status: waiting for inspection
Data ingestion from multiple sources
Telematics, gate logs, driver apps, ERPs and messages plug in as sources and feed one monitoring stream.
07:0014:00
Range for 10:00 updated · 33–62 min
Brain-like dynamic 'on the fly' learning
The model keeps learning what normal looks like from your own operation, and adjusts as it changes.
- Pickup dwell · Jebel Ali T244 min
- Transit · Jebel Ali → KEZAD1h 38m
- Reefer temperature · TRK-154.1 °C
- Idle time · Mussafah yard52 min
Real-time monitoring
The incoming stream is checked continuously, so every KPI shows where it stands right now.
- TRIP000212Route · E11On plan
- TRIP000213Route · E311On plan
- TRIP000209Route · E311+14 km off route
- TRIP000214Route · E11On plan
Automated detection of anomalous behavior
Readings that fall outside the learned pattern are flagged as exceptions before the process breaks down.
- New KPI · Pickup dwellLive from next reading
Measure
Gate-in → gate-out
Scope
Jebel Ali T2 · all transporters
Range
Learned per hour
Alert
Dispatcher · WhatsApp
Defining KPIs dynamically
Set up a new KPI from data already in the stream, and it is monitored from the next reading onwards.
Key benefits
Why teams switch it on.
Strong cross-check pipeline
An exception is checked against the other sources before anyone is disturbed.
Swift detection of anomalies
Abnormal readings are caught as they arrive, not in the next day's report.
Human-free KPI(s) management
Normal ranges are learned and kept up to date by the model, not maintained by hand.
Automated Alert(s) management
Alerts go to the right person, escalate on their own, and close when the KPI recovers.
Also in track
Milestones update themselves, exceptions escalate while there is still time to act, and the paperwork is captured as it happens.
Bring us one lane. We will show you the difference on your own numbers.
A week of your orders, rates or invoices, run through Fero with your contracts and your carriers.