Route planning: using algorithms to make the most of data
Smart city apps arrive in the home A medium-sized town somewhere in Germany, where residents separate their rubbish into glass, paper, food waste, plastic, old clothes and various other categories. All the bins are fitted with sensors to measure the fill level, and all the data is transmitted to a central server. We're talking here about thousands of bins. Millions, even, in large urban areas. Reliably transmitting signals, day in, day out. But who's supposed to process this information? The sheer amount of data generated would be way too much for even the brainiest human being to cope with. Big data needs smart software Even if we're not quite at the stage described in the scenario above, Binando already has the technology to make it happen. It only makes sense to collect data for the waste management industry, however, when this data can be turned into intelligent solutions, calculated using the right software. The solution which Binando has developed promises to help companies identify the shortest and most efficient routes for their waste collection vehicles. This is good for the environment, reduces congestion on the roads, and saves time and money. Accurate digital cartographical data is required for optimum route planning Artificial intelligence to improve urban living Binando has developed a complex algorithm which takes account of a number of factors. These include not just the actual fill level of bins or recycling containers, but predictions about when they are likely to need emptying. At the moment, the predictions are based on data collected in the past, but machine learning technology means they are set to become more and more precise as time goes on. On top of this, the system is able to prioritise individual containers according to their location, allowing bins in busy inner-city areas to be emptied more often than recycling containers out in the middle of nowhere. Once this has been factored in, it becomes clear exactly which bins need to be visited on any given day. Another thing to be taken into account is the fact that not every waste collection vehicle is suitable for every location. Imagine a huge rubbish lorry stuck in the narrow streets of a historical town centre, and the chaos that would cause. Binando's data processing software ensures that this kind of fiasco can never happen. The Binando app displays the planned route inside the vehicle, so the driver knows exactly where to go - all day long. Unbeatable: when humans team up with machines Once the software has identified which containers need to be emptied next, it begins to calculate the best possible route. Taking into account the number of vehicles available, the location of each and the amount of waste each can carry, the algorithm directs them round the relevant bins and then shows them the shortest route to their ultimate destination, e.g. the rubbish tip or landfill site. The software uses metaheuristics to generate the best course for the vehicles by checking, comparing and constantly optimizing all the routes found. The calculation process is stopped at a pre-defined point and the route last calculated is deemed to be the best. This proposal is then passed on to the planners at the waste management company. In the light of the volume of data likely to be generated once the system is fully in place, it is obvious that no human being will be capable of calculating and optimising routes with anything like the accuracy of the algorithm. This is not to say that human planners will become superfluous, however. They will still have to check and approve the routes generated automatically - here as in so many other areas, the best results are achieved when man and machine work together.