HKU Bulletin May 2008 (Vol. 9 No. 2)

6 Pinpointing Traffic Hot Zones The dangers on our roads are revealed thanks to innovative data analysis. O ne of the University’s leading researchers in transport development has established a new methodology for assessing dangerous stretches of road across Hong Kong. Dr Becky Loo Pui Ying, Associate Professor in the Department of Geography, is researching methods of identifying dangerous road sections. “I deal with the basic ideology that they are not accidents, they are crashes,” she said. “Accidents are purely random events, they cannot be avoided, whereas crashes can.” Starting from this viewpoint she and her team identified Hong Kong’s most dangerous stretches of road. “The government has identified black sites, or dangerous areas, and most of these are all in the urban parts, which makes the roads in the New Territories appear safer, but we just have to read the newspaper to know this is not the case. We know that areas like Yuen Long and Tuen Mun are quite dangerous locations,” she said. So, Loo went to work on plotting all the crashes in Hong Kong and found that they were spread all over the territory with clustering in different parts of the city. “I reduced these dangerous areas to 100 metre units to map the density of crashes and found the highest density in the Yuen Long, Tuen Mun, Sha Tin and Tsuen Kwan O areas. This type of analyses gives us a better idea of where the crashes are. But these are districts with no government-designated black sites at all. This is because the black site methodology focuses only on the road junctions and the 70 metre stretch either side. “If there are more than six pedestrian accidents in a year, or nine vehicle to vehicle crashes, it is a black site. Highways like Tuen Mun Road do not have many junctions, that’s why much of the New Territories appears to be safe, but they are dangerous and they are being ignored,” said Loo. The more sophisticated ‘black spot’ methodology deals with both junction and non-junction areas. Loo went on to dissect the entire Hong Kong network – 4,000 kilometres of road – into similar 100 metre units. She ended up with more than 59,000 road segments, in 18 districts. Then she assessed the actual crash rate of each unit, and compared urban with rural areas and established a critical crash rate. “If a unit exceeds the critical crash rate of, let’s say three crashes, we call it a hot spot.” A stretch of road with two or more contiguous hot spots becomes a hot zone. “A hot zone has six or more crashes because it consists of two or more units. So we have located 159 hot zones in Hong Kong, with a cluster in the New Territories, especially Yuen Long, Sha Tin and the Tsuen Kwan O areas. This is a vast improvement on the black site methodology,” said Loo. “A lot of our hot zones are on the expressways. So our methodology gives a clearer picture of where the hidden dangerous locations are. We can now prove that expressways in Hong Kong are very dangerous even though the existing black site methodology does not take them into account.” Loo is now sharing her findings with Government to help the Transport Department make more informed decisions when analyzing dangerous stretches of road. RESEARCH A Matter of Good of Taste One Professor has developed a new way of helping maintain quality control in the food tasting industry. A n electronic tongue has been developed in the Department of Chemistry that mimics the way humans taste and holds promise for better quality control in the food industry. The machine absorbs taste-related molecules and uses patterns to discern the type and concentration of flavour. Previous electronic taste machines could only measure the ions emitted by salty and sour flavours, but this one more closely imitates human taste processing and can also apply to bitter and sweet flavours. The machine was developed by Associate Professor Dr Fung Ying Sing with funding from the Innovation and Technology Fund (ITF). “Taste is very difficult for the food industry in terms of quality control. Different people have different sensitivities so even if you have a taste panel, they might not be able to detect tastes that some customers can detect,” he said. “Also, if the panel doesn’t agree, and there’s only one person who can make the final decision, what happens when they go? This kind of equipment can make quality control more consistent and reliable.” Fung and his team developed a molecular imprinted polymer for capturing, recognising and quantifying taste molecules. The information is fed into a computer, which uses patterns to determine the concentration and intensity of the flavour. In the laboratory he tested tonic water, which includes sweetener to mask the bitterness of quinine. The electronic tongue was able to distinguish both flavours. Word of the results has spread quickly and his team was called on by a Hong Kong orange juice manufacturer to test its product when it changed suppliers last summer. A machine with eight sensors was able to distinguish between orange juice and orange drink, and between drinks from different manufacturers. “It’s not yet subtle enough to detect differences in a drink produced by the same manufacturer. We need more sensors and more time to develop that,” Fung said. Ironically, the greatest interest in the electronic tongue is coming, not from companies interested in improving taste, but from those who want to eliminate it. Distilled water bottlers in the US, the Mainland and Hong Kong have expressed interest in the product to address customer complaints about a bad or mouldy taste in the water. Even when bottlers go through the expense of putting together a panel of tasters, they cannot always detect the problem. Fung is st i l l consider ing thei r request because this would require a different technological approach. Bad tastes in water are produced by algae and other sources, not taste molecules, and they would have to be measured and quantified using other mechanisms. “We’re still thinking about whether this could be an extension of the electronic tongue project,” he said. “In fact it is very different from our project. We have focused on developing taste, but drinking water shouldn’t have any taste.” 7