It has been a while since Kotihoito 2016, the national homecare conference in Finland, and one small but significant question from the audience remained in our thoughts after our presentation. During our presentation, a customer told how our optimisation solution, with automated long term work planning, had succeeded in creating a sense of stability in the usually chaotic setting of daily homecare; enough time is reserved for travelling from one place to another, the nurse has time to perform the planned client procedures and has still enough time for scheduled work breaks – All this with a 10% reduction in nurses working each day.
This was followed by a question from the audience; “How on earth did you manage to achieve this?”, and was asked by, no less, a representative from the Finnish nursing union, SuPer. Too often we hear very different stories about the practicality of homecare optimisation solutions which don’t live up to their promises, which is why this sounds too good to be true.
Science does not recognize the concept of miracles and engineers believe in them even less, so how was this achieved? In 2011-2012 we took part in the Vaske project, an extensive health care and social welfare development project for eastern Finland, where our logistics algorithms were tested for the first time in homecare planning. The project taught us a lot; the results looked excellent on paper, but in reality they were far from suitable for practical use. There might have been two nurses visiting the same building simultaneously attending to different needs of different clients and sometimes even the same client. Nurses were knocking on doors here, there and everywhere and one client could be visited by several different nurses during the same day. There was no similarity between days because the calculation was done on daily bases. If the objective was to minimize travel, the number of nurses per client increased too much. If having a personal nurse was emphasized, the nurses ended up travelling too much. The mathematics was not able to handle the number of variables in real life. Impossible, many would say, but fortunately our engineers like a challenge.
Five years later, after tens of thousands of hours of design, development, trial and error, including over 30 customer projects, what seemed impossible has become possible. It has taken countless mathematical algorithms, artificial intelligence, rules of thumb, hundreds of parameters, geography, optimized road map data and the fastest route calculation that ever existed. Could any engineer receive better recognition for their work than have someone state that they have achieved the impossible? We are now facing a new problem. Our sales department has a problem. Customers think our solution sounds too good to be true!