Project Title: Application of Heuristics to ordered Truck Loading Problem  
Student: Jonathan KING
Course: MSc in Computer Science (AI) 

Abstract:
Deliveries are a major factor in today’s modern-day life with small improvements 
to efficiency able to save companies significant money. This project looks at solving 
a real-life problem a company has at loading their truck whilst keeping multiple 
objectives and constraints in mind. This project proposes a constructive greedy 
heuristic to solve a multi-objective variable bin packing problem and multiway number 
partitioning problem in order to minimise space used, load balance of a vehicle and 
keeping orders of similar destinations together. The intention is to model the 
in a mixed-integer linear programming model as an experiment but solving the problem 
with heuristics to obtain solutions in a feasible time. This method is successful 
producing a set of pareto optimal solutions to any sized instance of the problem. 
Existing benchmark datasets are compared against along with the production of a new 
data set 10 times larger than the existing ones, available for future comparisons and 
testing on large scale instances.