icon-Support

Case studies

Our collaboration projects between the public sector, industry and academia bring a real impact to Scotland.

AI

Using Artificial Intelligence to design new logistics solutions

Forming part of the Highland and Islands Enterprise (HIE) led Northern Innovation Hub, the Highland Food and Drink Innovation Network was established in late 2019 with funding from the Inverness and Highland City Region Deal. Aiming to identify the areas where innovation would have the biggest impact on the sector, the Network carried out an engagement process with local Food & Drink producers, who reported that logistics was an area ripe for improvement.

The problem: Food & Drink producers in the Highlands and Islands often face demanding logistics requirements and challenging geographical conditions. These hurdles can make it more difficult and less cost effective to get their products to market and to receive vital supplies and raw materials.

The solution: Working in conjunction with a wide variety of partners, The Data Lab has helped apply artificial intelligence (AI) and nature-inspired computing to the logistical challenges faced by both logistics service providers and Food & Drink manufacturers operating in the region.

The challenge and the clear disadvantages for producers and suppliers

The Highlands and Islands of Scotland are home to world renowned Food & Drink producers, who play a key role in making Food & Drink the largest manufacturing sector in the UK. At the same time, these producers operate in the least densely populated area in Europe, often far from major transport hubs. Many producers are dealing in small volume shipments and frequently require special logistical considerations such as refrigerated storage or expedited delivery to the end consumer.

Servicing these remote producers, frequently accommodating such special requirements, can also be prohibitively costly for logistics suppliers. Inefficient routes and low volume shipments can lead to wastefully low utilisation of capacity in delivery vehicles, as well as unnecessarily high carbon emissions from road journeys.

Producers reported that the logistical challenges they faced resulted in significant inefficiencies and presented a barrier to entering or more fully exploiting markets elsewhere in Scotland, the UK and around the world. In addition, the lack of a centrally available source of information or coordination meant that no one producer, logistics service provider or trade association could take ownership of the issue. Finally, the complexity and scale of the logistics environment in the region meant that applying human expertise alone would not be sufficient to meet the challenge. The dynamic calculations involved in route planning, vehicle utilisation and data optimisation called for an AI powered solution.

The project

Rural Matters LLP, who were appointed to manage the Network, approached Highlands and Islands Enterprise to explore options for addressing the logistical challenges raised by members. Brought on board to provide expertise on the latest data driven solutions, The Data Lab was able to connect the Network with the Smart Data Technologies team at Robert Gordon University, as well as its commercial spin-out Celerum Ltd, who were already engaged in applying AI to large scale logistics planning.

Together, the parties designed a technical feasibility project funded under the Highland City Region Deal to investigate the question: ‘can Food & Drink logistics in the Highlands benefit from an open logistics software platform with AI capability to match food producers’ demand to move goods to logistics providers’ capacity to supply?

The three-month pilot project operated in two areas: Inverness and the Moray Firth, and the Isle of Skye. From their contacts across 190 member companies, the Network recruited 24 participants representing both Food & Drink producers and logistics companies operating in the pilot areas. The Network worked with the participants to gather information on their volume and frequency of logistics requirements as well as any seasonal fluctuations and special requirements.

Once this information was anonymised to protect the commercial sensitivity of all participants, it was sent to the University team to inform the AI modelling they hoped would identify more efficient logistics solutions.

Their approach is based on nature-inspired computing, which learns lessons from instances in the natural world where evolution has overcome inefficiency. The decision making processes which govern the flight paths of bees returning to their hives, for example, or the routes of ants collecting food, have evolved to eliminate inefficiency, and provide valuable lessons for human logistics planning. Using these principles to inform their algorithms, the team tested a variety of logistical strategies using the anonymised Network data.

Proof of concept to practical impact

The results of the initial three month pilot project were immensely encouraging. Applying models from the Smart Data Technologies team to the Network’s data identified the potential for significant economic benefits for both producers and logistics service providers. Furthermore, results indicate improved vehicle utilisation, which may result in carbon emission reductions of 20-40%, as well as increased resilience in the supply chain. Feedback from the study also indicated demand for the introduction of an open logistics software platform. This platform will allow producers and logistics service providers to place and receive orders in real time and use AI to plan optimised routes and vehicle utilisation based on the AI solutions tested in the pilot.

Having achieved proof of concept, the partners are now examining opportunities for wider testing and demonstration of the model. Further data collection from across the Highlands and Islands, from trade associations as well as from additional producers and logistics service providers, will help refine and expand the modelling. Consideration will also be given to the open logistics software platform itself, including what the platform might include and how it might be applied, operated and managed.

Those involved in the pilot project note that while it was developed in the Highlands, its findings and the resulting software platform may well be valuable to other regions with similarly dispersed small artisanal producers such as Italy and Spain.

The team hope that identifying more efficient logistics solutions will not only protect companies from previously vulnerable supply chains, but enable them to grow and reach new markets, protecting jobs and energising the local economies of Highland communities. The elimination of wasted capacity will also be a valuable step in achieving the carbon reduction necessary to meet the nation’s net zero targets.

Download the full report here: HiPLAN – Highland Food & Drink Open Logistics Platform 

Looking ahead

The Data Lab’s position at the cutting edge of applied data science in Scotland meant it was ideally placed to bring together the Network’s request and the expertise of Robert Gordon University and Celerum.

By providing funding, specialist advice and practical support, The Data Lab, in collaboration with key partners, supplied the momentum needed to apply data driven solutions to some of the most pressing commercial concerns in the region.

All parties now look forward to uncovering the full extent of the possible transformation as the project continues, and ultimately to delivering enhanced logistics solutions for organisations throughout the Highlands and Islands.

Find out more about the Data Drive HIE and how we could help your business.