• Neurolabs

A vision to serve, foodservice checkout reimagined

Updated: Jul 15, 2020


With growing levels of data, increasingly fast processing power and continuously improving algorithms, artificial intelligence (AI) solutions are now applied within an increasing multitude of industries across the globe. This is enabling faster decisions, actions and a stronger support base for countless operations day to day and making the most complex of issues simple to solve. As AI proceeds to be adopted across consumer, finance, healthcare, manufacturing, transportation and logistics, and governmental sectors, it is now time to focus attention upon the future impacts on foodservice, of which adoption of AI is still in infancy. Here we put forward the case for the application of cutting edge AI-enabled machine vision to the canteen checkout.

Customer pain points

For customers, it is a constant challenge to be served at a speed that does not drastically impede on their time. This is particularly prominent in higher traffic scenarios such as:

  • Busy staff canteens

  • Sport stadiums

  • Grab and go outlets

The average UK worker spends only 21 minutes on lunch. With 10–20% of this time spent queuing and checking out the time wasted daily is extensive.

The main problem is the checkout process. Today the alternatives to the traditional point of sale systems are pre-ordering apps and self-service kiosks used to pre-order items (such as the system in many fast-food chain restaurants).

Pre-order kiosks cannot be used to serve the whole crowd as there will still be queues to the kiosk itself which the use of a kiosk cannot speed up, merely moving the problem further away from the service counter. Mobile apps are also ineffective due to the struggles of user adoption, which many operators experience and are costly to develop, leaving a window for high-profit losses.

Operator pain points

Currently, the foodservice sector is facing the following challenges, impeding on the quality of service and profit:

  • Slow service rates (particularly in instances of high traffic) — On average the time taken to serve a customer via a cashier is 61 seconds. Add on queuing times and there is a significant impact on a customers experience.

  • Increasing operating costs — such as increased staff wages and rental costs

  • Customers that demand convenience — service culture weighs heavily towards speed over any other factor. If a customer can be served faster elsewhere, it is highly likely that they will choose to use this option.

  • An increasingly technology-focused culture

In order to provide solutions for customers and solve the issues of slow service, foodservice operators will need to look beyond the traditional standards of practice and adopt newer, more efficient practices. Solving these issues provides a clear and significant opportunity to positively impact huge numbers of people.

Assessing where value can be added, increasing checkout speed without compromising service quality is a crucial point for operators. We’ve created the solution to make this a reality.

Neurolabs’ AI solution (Zia)

A comprehensive solution to the issues of service in a high traffic foodservice environment, Zia provides a highly sophisticated checkout service that impacts further than simply providing a faster checkout.

Neurolabs’ solution provides this through:

  • Computer vision: the use of cameras in order to identify all objects presented simultaneously, with the use of AI algorithms. The solution is capable of going beyond the standard that competitors set, operating in poor conditions, with the ability to scan under inadequate lighting or in the instances when a product is partially covered from view.

  • Hardware: user-friendly and capable of processing high levels of customer traffic, achieving faster rates of service than a traditional point of sales system offerings. All through a self-service kiosk at the point of pain.

Benefits of Zia

  • A fully automated self-service system capable of serving 4x as many customers in the same time period as a traditional human-operated point of sale system.

  • A flexible system with the ability to increase the efficiency of serving in multiple scenarios, whether it be within a canteen-style environment, a cafe or even a food stand.

  • Reduction of space occupied by service and payment terminals previously inhabiting the end of the serving countertop. With Zia, outlets may be placed in any chosen spot leaving more space for customer service and stocking of products.

  • Zia will continue serving without any break in the flow of customers, meaning no more missed service due to a long waiting time causing fatigue of customer purchase intentions.

  • The ability to scan multiple items simultaneously means a reduced customer waiting time even when the customer’s chosen products are partially hidden from view.

  • Operators may now leverage staff time, made available through the implementation of Zia, to focus on furthering customer experience through other avenues such as serving food to more people simultaneously.

  • No missed transaction opportunities due to queuing fatigue, leading to up to 240 additional transactions per hour per checkout point at peak times.

Machine vision

The ability to scan multiple objects simultaneously through the use of cameras inbuilt into a point of sale system completely replaces the need for barcode scanners that identify products individually. The time contrast between multiple object identification and single object identification is extensive and leads to higher serving volumes (up to four times higher) within the operating day.

Synthetic data

The process of creating a database for a product is the combination of using 3D scanning and a Virtual Reality engine. The use of this engine enables Neurolabs to create countless scenarios in which synthetic data can be used to provide all the scenarios that Zia may encounter a product and generate labelled images for this context. This ensures the highest level of accuracy in identifying products (>98%), even in the extreme scenarios that the product is occluded, upside down or scanned in poor lighting conditions, or all three at once. Despite less than optimal conditions, Zia is still a fast and hyper-accurate solution for scanning and identifying products. In our testing, we have seen 2 seconds in comparison to an average of 46 seconds with a canteen cashier.


Given the challenges faced. Foodservice providers need to adopt a new technology in order to remain competitive and serve the increasingly demanding customer. A sophisticated and intelligent solution to checkout service problems, Zia allows you to revolutionise the experience of the customer, so that they may enjoy a seamless, streamlined service. Our vision to serve goes beyond the norm and seeks to empower users so that they may create the next generation of service.

If you would like a demonstration of the solution that we provide and see how we may positively impact your operations please contact us.

Contact hello@neurolabs.eu for further details.

32 views0 comments

Recent Posts

See All