You are using an older browser version. Please use a supported version for the best MSN experience.

Ailsa Land obituary

The Guardian logo The Guardian 14/06/2021 Georgina Ferry

You’re moving house and you want to pack the largest possible number of different sized items into a box, but don’t want the box to end up too heavy to lift. Or you have a series of related jobs to do, and you need to know what order to do them in so that you finish by a certain time. These are optimisation problems, and they were the kind of problem that fascinated Ailsa Land, who has died aged 93.

Before computers were widely available, Land devised pencil and paper methods to solve such problems without resorting to the “brute force” approach of trying every possible solution one by one. As soon as she could, she also grasped the opportunity computers offered to automate such methods, structuring data effectively and devising strategies to produce computer code that was as efficient as possible.

The “branch and bound” method that Land devised with her colleague Alison Doig (later Alison Harcourt) in 1960 “reshaped the landscape of mathematical programming”, in the words of the Operational Research Society. It is now a fixture of courses in maths, computing and statistics that have applications to real-world problems, and continues to be deployed in engineering, logistics and military operations.

Land pursued her interest in such problems in the course of an academic career that took her to a professorship at the London School of Economics in 1980, the first woman in the UK to hold such a post in operational research. Her early upbringing had made her fiercely self-reliant, and being one of very few female academic leaders at the LSE fazed her not at all.

Optimisation problems seized her interest early. As an assistant in the economics research division at the LSE in the 1950s, she worked on the well-known travelling salesman problem: how to plan the shortest route to make deliveries to a number of widely dispersed locations. She went on to undertake a PhD, using linear programming to analyse how best to transport coal from 154 British pits to 65 coke ovens, where it would be converted into coke for the steelmaking industry.

Always tackling problems from the perspective of an economist rather than that of an engineer, she arrived at a cost that was 10% lower than the National Coal Board had actually spent on transporting the fuel – a potential saving of some half a million pounds (more than £10m today) per year if extended to all UK pits and coke ovens.

The branch and bound method organised potential optimal solutions of a problem as a tree of branching possibilities

The branch and bound method was designed to cope with the particular case of problems that need solutions that are whole numbers, rather than continuous variables such as weights of coal or periods of time. It was a way of organising the potential optimal solutions as a tree of branching possibilities, which allowed you to stop pursuing a whole branch if it produced a solution that was less good than one you had already found on a previous branch. The work was funded by BP, to improve the linear programming they were using to run their refineries.

Land and Doig’s 1960 paper stimulated research in mathematical programming all over the world. She herself continued to experiment with branch and bound until after she retired, applying it to large travelling salesman problems with up to 150 locations to visit.

Once her unit at the LSE had ready access to computers, she devoted much of her energy to writing routines that could implement approaches to linear programming. With her colleague Susan Powell, she published Fortran Codes for Mathematical Programming in 1973, and she shared her code openly for the benefit of colleagues.

She also took a lead in teaching, setting up new courses at undergraduate and master’s levels and mentoring a succession of PhD students. Her colleagues found her unfailingly encouraging and optimistic, gracious in manner and adept at avoiding “institutional entanglements” in her leadership of the operational research group.

Her continued influence on the field was recognised by awards throughout her life, most recently by the Operational Research Society’s Beale medal in 2019 and, posthumously, the gold medal of the Association of European Operational Research Societies in 2021.

Born in West Bromwich, Staffordshire (now in the West Midlands), Ailsa was the only daughter of Elizabeth (nee Greig) and Harold Dicken. Her father worked in his family sports retail business and was later a salesman for Dunlop. Ailsa developed a passion for science, but as she chafed at the discipline of her local grammar school in Lichfield, her parents sent her to a small, mixed boarding school, Rocklands, in Hastings, East Sussex.

She had been there only a year when her mother swept her off to Canada, hoping to spend the second world war with relatives there. After moving several times they settled in Toronto, where for three years Ailsa went to the Malvern Collegiate Institute (though she was still in her early teens, it was the seventh school she had attended).

When Canada began to recruit soldiers, in 1943 both Ailsa and her mother decided to join the Canadian Women’s Army Corps (Ailsa claiming to be 18 while two years younger) and by 1944 found themselves in clerical jobs in the National Defence Headquarters in Ottawa. That year they obtained compassionate discharges to return to the UK as Harold, who was serving as a catering officer in the RAF, was undergoing a dangerous operation (which to their relief he survived).

As a demobilised servicewoman Ailsa was able to enter the LSE to study for a degree in economics in 1946: she graduated in 1950 and spent the rest of her career at the institution. Among her graduating class she met Frank Land, who had come to Britain with his parents and twin brother in 1939 as refugees from Nazi Germany. He went on to become one of the computing pioneers who developed the Leo computer for J Lyons & Co, and is also an emeritus professor at LSE. They married in 1953 and had three children during her PhD studies: true to her calling, Land managed her time optimally so that she could enjoy both family and career.

Following her retirement in 1987 from teaching and administration, she continued to work on research projects, only slowing down when she and Frank moved to Devon in 2000. She enjoyed adventurous travel, including trips to Australia, New Zealand and the Antarctic, and engaged with village life as clerk to the parish meeting in Harford, near Ivybridge. From 2015 they lived in Totnes, spending more time with their growing family.

Ailsa is survived by Frank, and by their children, Frances, Richard and Margi, seven grandchildren and two great-grandchildren.

• Ailsa Horton Land, professor of operational research, born 14 June 1927; died 16 May 2021


More from The Guardian

image beaconimage beaconimage beacon