AFS: Improving how we identify fleet damage using AI technology

Last-mile delivery companies like AFS and their drivers are under increasing pressure to deliver goods fast and within a specific time frame. It’s estimated that delivery drivers in the UK drop off a parcel every six minutes on average.

With this additional pressure, even the safest and most competent drivers, are frequently sustaining minor damage to their vehicles – particularly in towns and cities where streets are narrow, and parking is limited.

What’s the cost of this damage?

The financial impact of these small bumps and scrapes is spiraling out of control for business owners, most of which goes unreported by drivers.

To fully understand the extent of this damage, AFS carried out an extensive analysis of damage to its fleet (c.300 vehicles) from November 2021 to October 2022. The figures were an eye-opener…

Over an 11-month period, AFS amassed £440K of damage to its fleet, with £115K in unreported damage alone. Perhaps even more jaw-dropping is that out of 1515 vehicle damage points recorded during this time, only 6.9% (145 damage points) had been reported by drivers/managers.

Why is reported vehicle damage so low?

Even with the rigorous vehicle checking processes, AFS has in place to help identify damage –vehicle inspection photos, monthly van checks, vehicle damage auditing, and vehicle un-assignments, a lot of damage still goes under the radar.

Unless third parties are involved, or the damage is significant, low-impact damage can often go unnoticed, particularly during peak season when pressure on drivers and managers is immense. Additionally, drivers can sometimes be reluctant to report damage, fearing a loss of work or incurring the cost to repair the damage.

Why do we need to keep on top of the damage to our fleet?

Aside from the financial impact on the business, our ability to be able to quantify and identify damage to our fleet has many benefits, including:

• Keeping drivers safe
• Protecting vehicles
• Better planning for repairs and vehicle downtimes
• Safeguarding drivers from inheriting vehicle damage debt they didn’t cause
• Helping drivers with affordable payment plans to cover the cost of damages

In addition, at AFS we reward safe drivers, which creates a positive working environment and helps ensure we retain talent in our team.

Using the power of AI to improve damage identification

After reviewing the results of the fleet damage analysis at the end of 2022, AFS knew something had to be done to help improve how we can quickly identify damage. And so, we took the decision to trial AI technology to analyse vehicle inspection photos.

Over a five-month period, our team carried out almost 2,000 inspections using AI, manually checking the results against our vehicle schematics. The accuracy of the AI technology was recorded at an impressive 92%!

How does the AI technology work?

• All drivers and managers have an app on their smartphones that they use to scan the vehicle and take photos.
• The app automatically identifies all panels on a vehicle and uploads these images into the AFS platform.
• These images are analysed for new damage and results are available typically in around 30 minutes.
• Managers at AFS are alerted to any new damage.

 

“The 5 month trial of the AI technology was very labour intensive – 12,000 photos all manually cross-checked by our team. But the results were even better than we were hoping for! Going forward we know that by using this technology, the time and money we’ll save by identifying damage to our fleet will be significant.”

– Kieran Hughes, Fleet Manager at AFS.

What’s next for AFS?

Following the success of the 5-month AI technology trial, AFS has taken the decision to adopt the AI technology permanently and plans to use it regularly to analyse vehicle inspection photos. It’s estimated that the time alone saved by AI technology over manual photo inspections is the equivalent of a full-time job, not to mention the other benefits of early identification of damage to vehicles.

In addition, the insights mined from the trial have been invaluable. In particular, the data has highlighted that most of our damage occurs from low-impact, maneuvering/reversing incidents. We now have an opportunity to better understand why these incidents occur and how we can effectively manage and reduce these types of incidents.

Interested in using AI technology to manage damage to your fleet?

The results of the trial of this technology speak for themselves. And by testing out this technology on our own fleet at AFS, we have now rolled out the feature into our UniQore software platform, so other last-mile delivery businesses can benefit too.

Is this something you would be interested in for your business? To find out more about how UniQore and AI technology can benefit your business, get in touch for a chat.