On the first day of the May Day holiday I stayed in the office and worked through the 2022 data center question. After three hours I had read the brief and produced a scheme, but when I started the formal draft, half of it felt wrong. I wanted to adjust it but had no clear direction. If a plan is not scientific, it is not reasonable. So I found a reference solution, and the fog lifted at once. I could see my own mistakes and felt I had learned something, but being wise after the event will not get me through the exam. I need to use the coming practice sessions to improve enough to pass on May 16.
Where the scheme went wrong
The first problem is still the limits of my knowledge and practical experience. I am not sensitive enough to the clues that unlock a question. When I saw an open stair, my first thought was not that it was a straight open stair; I was still locked into the double-run corner-stair pattern. With repeated practice, certain keywords should immediately tell me what shape a room must take. I am also not sensitive enough to alignment relationships. In principle, if a room has a fixed and exact dimension, its position can be used to derive the dimensions of rooms along the vertical and horizontal directions. I have not mastered that. It is why half of my scheme was right and half was wrong. I failed to use the correct half to derive the incorrect half, so the solution got stuck.
Still, these practice sessions have improved my planning. I now know that some rooms must be analyzed as a group, and some have alignment relationships. I also know that when a room area seems to be missing, there must be circulation space that combines with it to make a standard area figure.
The second problem is that my planning instincts are too small. I saw that the scheme needed a circulation corridor, calculated the area, and decided on a 2 m corridor. The reference solution used 4 m. My thinking was too narrow. The root cause is that I still lack a designer's mindset: I am afraid to commit, pinching every meter like someone who has been poor for a long time. That habit needs repeated practice. When I dealt with the security inspection hall and the exhibition area lobby, I had already worked out that they should be analyzed together and that a stair should connect next to the exhibition hall. But then I made the security hall square, which was far too cramped. From ordinary experience, security halls are rectangular. Think of a high-speed rail station or an airport terminal: the room is long so more gates can be installed and people can move in and out. Yet when I sat down to draw, I treated it as a square. That was a serious mistake.
The third problem is that I did not read the question carefully. Careless reading is another root cause of poor planning and weak problem-solving. I did not fully absorb the meaning of the text. For example, I handled the science exhibition hall, multimedia hall, and digital application experience hall incorrectly because I read the requirement that each side should be no less than 12 m as if only one side had to be no less than 12 m. That was actually a good breakthrough point. No less than means the side must be 12 m. Divide the area by 12 and you get the other side. Combined with the column grid, you can also work backward to a corridor width of 4 m. Because I misunderstood, I chose a 2 m corridor and calculated a wrong answer. Once the answer was wrong, combining rooms from it could never produce a regular plan.
I also ignored parallel relationships in the bubble diagram on my first reading. My first arrangement put the science exhibition hall and the other two halls in a series, and I did not see the circulation line running alongside them. In the same way, entering from the data machine room area lobby, the bubble diagram clearly showed that the lobby connected the supercomputing room, management room, and lounge; the supercomputing room and the lounge even had two circulation lines between them. Simply put, no line on a bubble diagram is drawn without reason. Every detail is a key to the question, and it has to be analyzed carefully.
What I need to fix
Variable bays are my first big problem in solving these questions. Ever since the 2018 bus hub passenger station project, variable bays have troubled me. Whenever the suggested column grid was 8 m, I would force the whole solution onto an 8 m grid and never consider that the bay could be adjusted. There are two reasons: I lack design experience and sensitivity, and I read the question carelessly without understanding what lies behind it. Both make me timid and unwilling to test alternatives, or else the first reaction never occurs to me. A variable bay is not arbitrary; it follows the design logic.
Alignment is not only about grouped rooms and rooms above and below each other. It is also about the alignment of the whole scheme. For example, once the bay width of a certain room is fixed, I can use it to align rooms in that direction and infer what should be placed there. If I can find a matching room, I can also work backward to check whether the room I have drawn is correct and reasonable.
Rationality matters too. A building is for people, so from a human point of view a correct scheme must be scientific and reasonable. A rectangular security hall is more reasonable than a square one. When solving a question, if I hit a problem I cannot resolve, I should ask what arrangement would be scientific. That can help me derive the scheme quickly. In the exam, if I lack sensitivity, my brain seems to short-circuit and I cannot think of the answer for a while. Then I can only keep testing until I reach a scientific result. At that point I need to borrow from real cases, use rooms of the same type to reason forward, and push my thinking beyond what I usually consider.
Experience is another weak point. Many building types are unfamiliar to me, or I do not remember them clearly, so I cannot think of the answer quickly. Unless there is no other choice, I must not use divergent trial and error; it wastes precious time. In normal practice I need to sketch more small plans and draw common combinations of lobbies, stairs, elevators, and toilets. I do not need muscle memory, but I do need a strong impression, so that when I see something in the exam I can immediately think of how it should be combined.
A correct scheme must also be complete. Completeness means the spatial shape is complete, with no missing corners, no scrambled circulation, and no required room left without a place. It also means the building has a degree of symmetry. In this design, for example, from the exhibition area lobby one faces the courtyard, and the visitor passage hall and office rooms behind it sit on the same axis. A complete scheme starts with complete circulation: straight, orthogonal lines divide the plan cleanly, keep functional zones intact, and avoid uneven, jagged edges.
A better way to read the question
When the exam paper is handed out, I should use the completeness check to look at the design task and any special legends, then glance at the site plan to form an initial impression. In the few minutes before the exam begins, I can run through it mentally.
In this question the design task was a data application center. The special legends showed plans for two machine rooms, which meant the machine-room problem was central. In 2018 the legends showed parking spaces and an escalator, which meant the escalator had an effect on the design. Any legend that is given must be drawn exactly as shown.
The complexity of the site plan also affects the exam. If the site plan contains many elements, the solution cannot be separated from it; the breakthrough will come from combining the two. In a comprehensive-building question with an existing office building, for example, the connection between old and new buildings will affect functional zoning and circulation. If the site plan is simple, as in this task, with little beyond the site boundary and surrounding roads and not a single building, the main difficulty lies in the design itself. In that case, after checking the north arrow I can be bold.
When I begin reading in earnest, I should first look at the other requirements and the drafting requirements. The drafting requirements tell me what I must draw, so I can sketch a simple site plan and mark the key items: how many site entrances and exits, whether a ring road is needed, how many parking areas, and so on. I should also mark important items on the question paper so I do not omit them when drawing.
The most important piece of information in the other requirements is the basic column grid. In this exam the grid was 8 m, so room areas had to relate to 8: 64, 48, 24, and similar numbers. Once I know the grid, I become more sensitive when reading the area table. I can recognize the common related areas and see how they combine into clean rooms.
Next I should read the site plan design requirements. They give the sizes of parking areas, roads, and assembly spaces, and the positions of entrances and exits. I should mark and draw these directly on the site sketch. Knowing the entrances to internal and external roads helps me identify internal and external functional zones and gives the rooms in each zone a point of reference.
At the same time I should use the bubble diagram. It clearly labels the names and number of building entrances. I can combine that with the internal and external zones and make a short list, so I do not omit anything when drawing the site plan, and so the next stage of reading and solving is easier.
The architectural design requirements are the most important part, worth about 80 points. If I handle this well, failing becomes difficult. This section must be read closely and carefully. I failed to solve today's scheme because I read carelessly. In past years I failed because I did not understand the meaning of the text. It really is necessary to go word by word and analyze what lies behind the words.
This step must revolve around the bubble diagram. I should combine it with the area table and complete the bubble diagram while reading. The architectural design requirements are organized by zone, and the area table shows at a glance how many rooms each zone has and what each room requires: area, whether it needs ventilation, and so on. Some rooms in the area table may not appear on the bubble diagram, so I need to add them, along with their area, whether they need daylight and ventilation, and how many grid cells they occupy. If the text gives suggested length and width dimensions, I must adopt them and draw them prominently on the sketch so I do not forget after reading the whole question.
Today I ignored the requirement that each side of the science exhibition hall and the other two halls should be no less than 12 m. One mistake led to many more. To prevent omissions and misunderstandings, I should read this part at least twice: once normally, then again as a check, and I should verify that my notes match the original meaning.
I also need to study where the given legends must be used. The shape and size of a legend has a strong effect on room layout. Today's auxiliary machine room plan directly fixed the relationship among the air-conditioning equipment room, the auxiliary machine room, and the visitor corridor. That was a good breakthrough point. In the 2019 multi-hall cinema question, the layout of the cinema auditoriums was given. If I cannot use that properly, the question cannot be solved.
After reading, I arrange the rooms and draw the sketch. If I have fully understood the question, it is easy to produce a scheme, but the easier a scheme is to draw, the less likely it is to pass. The large design question is the hardest subject in the exam, a touchstone and a roadblock. If the tiger lets you take it down easily, either you are an expert or you have done it wrong. Every year some candidates finish their scheme and feel satisfied, then fail. Others think their answer is barely different from the standard solution and expect a high score, yet they fail. That means some detail in the requirements was missed.
How I should check my own scheme
How can I tell whether my scheme will pass? The standard seems simple. First, when I look at it, it should feel good and give me confidence. Today my first sketch of the second floor satisfied me, but when I reviewed the first floor, something felt wrong everywhere. In the shared work area, I first drew a courtyard-shaped ring, then changed it to a single bar while working on the second floor because I thought a single bar would also work well downstairs. The standard solution used the ring. I had changed a correct move into a wrong one.
Second, after I am satisfied, I should look at the scheme from a human point of view and see whether it follows the underlying logic of design. Are the rooms usable? Making the security inspection hall square was clearly bad; when crowds arrive, people would jam at the entrance. Is the circulation smooth? If traffic on the same side has to twist and turn, the scheme cannot be scientific. Are the rooms regular? No matter what, a room should not be irregular. If I draw an irregular room, I should give myself a slap.
Alignment between upper and lower rooms also matters. A friend of mine, practicing the multi-hall cinema question, put toilets in different positions on two floors. He told me that a teacher had said toilets could be placed in different positions and did not have to stack vertically. I said that even if it only cost a few points, such an unscientific arrangement would definitely affect other rooms, and in that situation passing is basically impossible.
A qualified scheme should be pleasant to look at: smooth circulation, clear zoning, reasonable room arrangement, and an overall sense of symmetry and order.
Tomorrow I plan to do two sets of ink-line drawings to practice drafting. It will also give me a breather. The day after, I will try another real question and see whether I have improved. I hope I can pass the exam.