AI analysis of free text makes for smarter logistics

A major logistics company had a problem that was causing time and money to be lost as well as customer service impacted. Vehicles were being sent out to pick-up consignments for delivery only to discover on arrival that there was some kind of special requirement. The items might be particularly heavy or difficult to handle, and thus require a tail-lift or other specialist equipment, for example. It was often the case however, that the client had mentioned the special requirement – but usually, this would be written into the free text field on the online request form. This would not always be picked up by agents in the company’s offices and subsequently, vehicles that were unsuitable or did not have the appropriate equipment would be dispatched. To counter this, SCC devised a system that would use natural language understanding to analyse the free text input-fields on delivery instructions and determine if specific requirements need to be met for that job. To make this work efficiently, we analysed a sample set of around 14,000 free text entries and trained the algorithms to understand certain key criteria, such as when a truck with a heavier load capacity or a tail-lift would be required, for example. By doing this we were able to identify the hidden delivery criteria in more than 95% of cases. As a result, the logistics company has been able to make better use of its available resources and reducing cost. Most important of all, the system is helping to improve customer satisfaction as the company is able to send the right vehicle to each job, first time. As a result, we’d also expect there to be a positive impact on the company’s reputation and net promoter score. As well as natural language understanding, we have deployed AI and analytics technologies here. The resulting solution is effectively a simplified form of robotic process automation. Without these technologies, a company with any similar challenge would have been compelled to change its process or add an extra level of manual checks – and with it, further complexity. This would have had cost implications and slowed down processing of orders, with a resultant impact on customer service and productivity. Perhaps even more significantly, it would still have been prone to human error and interruption by unforeseen circumstances, and thus much less effective as the solution we have been able to deliver. Click here to read how emerging technologies can be used in your organisation. For more information on our Innovation services please click here.