June 2019 Summaries

On-demand last mile transportation: Real-time route optimization with Location Intelligence` is a complex problem that requires advanced algorithms and techniques to solve efficiently. The traditional last mile transportation problem is a vehicle routing problem (VRP), which is a combinatorial optimization problem known for its complexity from an algorithmic design point of view as well as computationally. On the other hand, real-time on-demand transportation problems are smaller problems because there is less information available but a response is needed immediately. The main challenge in solving this problem is to calculate all the required costs and design an algorithm that can find a near-optimal solution in a matter of seconds. Traditional logistics solutions cannot be adapted to on-demand transportation problems due to their complexity. To overcome this, spatial data science can make a difference by building data models that simulate existing conditions providing insights on existing constraints inefficient assignments and much more. Two initial approaches to solve the problem are the greedy algorithm and the batch assignment algorithm. The greedy algorithm searches for the nearest idle rider and assigns the order to that rider, while the batch assignment algorithm postpones the order assignment and runs the algorithm every n minutes. The batch assignment algorithm improves the assignments by having more information available at each iteration of the problem. To further improve the assignments, broadening the information available at each iteration is essential, such as considering what could happen in the following 30 to 60 minutes or even more time depending on the length of the services. Other improvements that would lead to more efficient and higher quality assignments include optimization criteria, such as time, utilization, customer and rider experience, matching the quality of the customer in terms of the frequency of use of the service and the quality of the rider based on the customer’s opinions and efficiency in delivery, combining orders, and creating larger batches by postponing the assignment as much as possible.

The UK has experienced persistent socio-economic inequalities over the past forty years, with the Gini coefficient rising from 0.27 in the 1970s to 0.52 in 2010, indicating a widening gap between the top and bottom 20% of the population. The UK2070 Commission's first report highlighted regional imbalances in the economy, prompting a formal inquiry into these issues. A project at the University of Liverpool's Geographic Data Science Lab explored "Trajectories of Neighbourhood Change in Britain" using census data to create geo-demographic classifications and interactive digital maps. The research revealed that neighbourhood trajectories have shifted over time, with some areas experiencing upward thriving, increasing socioeconomic diversity, or struggling home-owners, while others are stable affluent or multicultural urban. The project aims to improve prediction of inequality and explore the sources of this issue using machine learning models and satellite imagery, ultimately helping policymakers address inequality at its source.

The utilities industry is facing significant challenges, including updating grid infrastructure and responding to a rapidly evolving regulatory landscape. Innovative companies are exploring the use of Artificial Intelligence and IoT to optimize business models and meet changing customer expectations, with Spatial Data Science playing a crucial role in driving success. Many utility providers intend to ramp up their investment in geospatial capabilities, but face challenges in attracting top talent due to high demand from tech giants and financial services. Geospatial data science is being used to address various business problems, including EV infrastructure rollout, solar potential analysis, and churn reduction. For example, spatial analysis can help utilities companies understand the location of electric vehicle buyers, optimize charging infrastructure placement, and identify areas with high churn rates. By leveraging location intelligence, utilities providers can increase occupancy profitability, customer satisfaction, and revenue generation from public charging points and home charging services.