(1) Paper Title A Multi-Objective Genetic Type-2 Fuzzy Logic Based System for Mobile Field Workforce Area Optimization (2) Author Contact Details: Andrew Starkey, CSEE Dept. University of Essex, Wivenhoe Park, Colchester, CO4 3SQ, United Kingdom, astark@essex.ac.uk (+44)7581183308 Hani Hagras, CSEE Dept. University of Essex, Wivenhoe Park, Colchester, CO4 3SQ, United Kingdom, hani@essex.ac.uk Sid Shakya, Business Modelling and Operational Transformation Practice, British Telecom, Adastral Park, Ipswich, United Kingdom sid.shakya@bt.com Gilbert Owusu, Business Modelling and Operational Transformation Practice, British Telecom, Adastral Park, Ipswich, United Kingdom gilbert.owusu@bt.com (3) Corresponding Author: Andrew Starkey astark@essex.ac.uk (4) Paper Abstract: In industries which employ large numbers of mobile field engineers (resources), there is a need to optimize the task allocation process. This particularly applies to utilities companies such as electricity, gas and water suppliers as well as telecommunications. The process of allocating tasks to engineers involves finding the optimum area for each engineer to operate within where the locations available to the engineers depends on the work area she/he is assigned to. This particular process is termed as Work Area Optimisation and it is a sub-domain of Workforce Optimization. The optimization of resource scheduling, specifically the work area in this instance, in large businesses can have a noticeable impact on business costs, revenues and customer satisfaction. In previous attempts to tackle workforce optimization in real world scenarios, single objective optimization algorithms employing crisp logic were employed. The problem is that there are usually many objectives that need to be satisfied and hence multi-objective based optimization methods will be more suitable. Type-2 fuzzy logic systems could also be employed as they are able to handle the high level of uncertainties associated with the dynamic and changing real world workforce optimization and scheduling problems. This paper presents a novel multi-objective genetic type-2 Fuzzy Logic based system for mobile field workforce area optimisation, which was employed in real world scheduling problems. This systems had to overcome challenges, like how working areas were constructed, how teams were generated for each new area and how to realistically evaluate the newly suggested working areas. These problems were overcome by a novel neighbourhood based clustering algorithm, sorting team members by skill, location and effect, and by creating an evaluation simulation that could accurately assess working areas by simulating one days worth of work, for each engineer in the working area, while taking into account uncertainties. The results show strong improvements when the proposed system was applied to the work area optimization problem, compared to the heuristic or type-1 single objective optimization of the work area. Such optimization improvements of the working areas will result in better utilization of the mobile field workforce in utilities and telecommunications companies. (5) Relevant Criteria: -D, The result is publishable in its own right as a new scientific result independent of the fact that the result was mechanically created. -E, The result is equal to or better than the most recent human-created solution to a long-standing problem for which there has been a succession of increasingly better human-created solutions. -G, The result solves a problem of indisputable difficulty in its field. (6) Statement of Why Criteria Has Been Met: D: The system solved a complex real world problem using genetic algorithms. The results generated from this system had a real impact when put into operations. It produced optimal geographical structure for engineers to operate, by grouping 5400 working exchanges in BT into 250 working areas, such that the travel, coverage and utilisation of technicians is maximised. We combined genetic algorithm with clustering technique to reduce the search space, so that the evolutional only needs to find the center exchange building of the area rather than finding the configuration of exchanges for all 250 working area. In addition to this we restrict the possible designs to only ones that are logical for our problem by meeting our constraints. For example we created a neighbourhood clustering algorithm that will only connect points if they share a geographical border. So the decision to add a point to a cluster is not purely based on distance. Any distances we do use to make decisions, can be handled by our fuzzy systems. E: Human created solutions to this problem are sub-optimal in their design. This is because the number of design options increases exponentially with the number of exchanges. Additionally, humans are far slower, taking weeks or months to produce their results, whereas the machine option can take minutes or hours. Humans created design were previously preferred as they could factor in local knowledge of the area and how resources interact with each other. Any automated approach were limited to rule based solution which would not be able to match the quality of design produced by human due to complex rules involved in handling design constraints. However use of evolutionary algorithm have allowed us to incorporate human knowledge into simple objective functions and evolution process handled all the complexity of moving to better design. The resulting design is far superior than that produced by human, as evidenced by the savings given by the increase in productivity seen so far in the machine optimized areas. G: Given that the problem scales exponentially, even small companies that want to tackle workforce optimization will face difficulty without some machine intervention. If a company has just 90 locations to monitor that would lead to over 56 million ways to divide up the locations, ranging from 1 cluster to 90. When you have companies that have hundreds (or thousands) of locations and resources, the levels of information (and uncertainty in the information) grow to a level that is problematic and time consuming for a human to analyse without help from a machine. Given that this is a combinatorial problem meta-heuristics are a good option, in our case genetic algorithms. One reason for this choice is that we need a good solution in a reasonable time. Even if the solution isnt the most optimal the GA outperforms the human alternative in both these aspects, producing better solutions in minutes compared to months. (7) Citation: A. Starkey et al., A Multi-Objective Genetic Type-2 Fuzzy Logic Based System for Mobile Field Workforce Area optimization, Information Sciences 329 (2016) 390-411 Official Link: http://www.sciencedirect.com/science/article/pii/S0020025515006684 (8) Any prize money, if any, is to be divided equally among the co-authors. (9) Why The Judges Should Consider The Entry As "Best" In Comparison To Other Entries We have successfully used genetic algorithms to design working areas for BT engineers to operate within. The effects of the genetic algorithm designed working areas have been noticeable and have benefited 3 areas: 1. Operationally we have seen a 0.5% increase in productivity and an estimated 2 Million in savings. In addition the reduction in travel by the engineers over the same period was estimated to be 2.9%. This led to a saving of over 143,000 in fuel costs. The continued use of the system will mean more of the old human made designs get replaced by the genetic algorithm designs. This has the estimated impact of a 5% uplift across the UK. 2. Society has benefitted from the lower C02 emissions from the reduction in fuel consumption. This also helps the company meet sustainability targets. 3. Finally there has been a level of knowledge dissemination on this topic that will allow other companies to look at the successes we have had with genetic algorithms. These successes were highlighted to the telecoms community, when this genetic algorithm based optimization technique was recognised with a Global Telecoms Business award for business innovation. This will hopefully encourage them to come up with their own implementations to take advantage of the economic and environmental benefits. (10) Method used: Genetic Algorithms (GA).