There is an unfortunate side effect to being exposed to Lean thinking in large quantities: you begin to see so much waste in processes all around you. It is everywhere, and going Lean could help all these companies. That includes amusement parks!
I visited the Särkänniemi amusement park in Finland with my family today, and this is a tale of what my kaizen eyes (which are by no means perfect) saw on that trip. I will use Legoland Billund as a comparison a fair bit, as I visited that amusement park earlier this year.
Wristband tickets and hand stamps
I bought our tickets in advance from a partner retail store, from which I received the wristbands. The wristbands have a blank space where the customer needs to fill in the date of use. There are instructions on how the date should be entered on Särkänniemi website and on park maps that are available inside the amusement park (after the inspection point where you need to have filled in the date). No such instructions were given to me when making the purchase, but luckily the employees were not that strict about the date format. This, of course, raises the question why the format is carefully instructed in the first place.
The date information was checked sometimes carefully and sometimes not at all when entering rides. When it was checked, the wristband usually had to be rotated, which took additional time.
The wristbands came in three colors according to the height of the user. However, there were five height classes for admission to the rides, so users who were close to the limit got a stamp on their hand after the first height measurement that indicated which rides they were allowed to ride. There were several hand stamps in multiple colors, and the operators had them all in small boxes, from which they had to look for the correct stamp and the correct color. This took quite a lot of time each time someone needed to get their hand stamped. A minor improvement would have been to apply 5S to the stamp boxes so that the tools would have been arranged in the same fashion in each box and easily accessible at each ride. I am not quite convinced that the stamps are the way to continue though.
Wristbands and hand stamps are not part of what people pay for when visiting an amusement park, anyway. In Legoland Billund, for example, there are no wristbands: you pay for entry and are good to go from there.
Särkänniemi faces a small problem in this regard, as there are a couple of restaurants and an art museum in its area, and access to those is for a discounted fee that does not include the rides. They also effectively split the park in half, so they cannot be isolated into a separate area. Särkänniemi has also put in much effort to selling wristbands in advance in order to reduce queues at the main gate.
If Särkänniemi could do without the discounted fee for area access, it seems that all their goals (people of correct height have access to rides, no huge queues at the main gate) would be best served by making area access free and selling wristbands for all five height classes (for example, in different colors) in all stores in the park. This would greatly increase the number of sales points and simplify the wristband policy by eliminating stamps and hand-written dates. It would also remove the need for height measurement at rides altogether.
Another alternative for height measurement would be to do it as Legoland Billund does, by measuring the height at the rides for all borderline cases. With easily accessed measurement poles, it does not take a lot of time, but as Särkänniemi still needs a way to identify people entitled to access the rides, they cannot do away with the wristbands altogether, so this is not as ideal a solution there.
Organizing the queues
There is one height measurement related takeaway for Särkänniemi to learn from Legoland Billund though. At Legoland, there are height measurement poles near the beginning of the queueing area for each ride. At Särkänniemi, I observed people being turned away from rides by the operators multiple times. At least some of this useless queueing could be avoided by having clear indicators available at the point where the customer decides whether to queue or not.
Another inefficiency regarding queues was that there were no marked queueing areas beyond a very small area right in front of the ride. Queues often continued way past the marked area, which caused confusion as to who exactly is in the queue and to which direction does the queue extend. Simple cordons could provide temporary extended queueing areas and be easy to remove from use when there are no significant queues.
Yet another inefficiency regarding queues was that there were no indicators as to the expected queueing time. Because the capacities of the rides vary a lot, it is impossible to judge how long it takes to queue to a particular ride without an indicator, and this prevents customers from making informed decisions on where they wish to queue. In Legoland Billund, the operator of each ride had a manual board with queue length cards at five minute intervals. Determining the queue length based on where the end of the queue is should be relatively easy when the capacity of the ride is known. Displaying this information could help level the demand as customers could easily see how long it takes to get to a specific ride.
One of the key ideas of the SMED (single-minute exchange of die) technique, which attempts to reduce the setup time in machining, is to move any setup you can to the time period when the machine is running as opposed to a separate setup period. Legoland Billund had done this rather nicely in their Dragon roller coaster and Mine Train: customers were pre-assigned to specific carts by splitting the queue to cart-specific sections at the end. This reduced waste from setup while still enabling capacity to be fully utilized. This method would be perfectly applicable to Vauhtimato, which is one of the rides with the longest queues in Särkänniemi. All it would take is changing the end of the queueing area a little.
Rides and process waste
Now we get to a point that is very characteristic of Lean, analyzing the non-value added time in processes. In this regard, there is really no difference between amusement park rides and machine tools in a machine shop.
Customers want to attend the rides. Thus, the rides only add value when they are serving customers. Setup time (in this case, the time it takes to move the old passengers away and the new ones in) is waste.
Each ride has an ideal capacity, which is the product of the number of seats and run time. For example, an 18-seater ride with a run time of 2 minutes could ideally serve 540 customer per hour. Of course, this ideal can never be reached, as we cannot teleport people to and from the ride, but Lean is about striving towards this perfect state nonetheless.
I did not conduct extensive testing, but I did grab my stopwatch for a while to get some data. For example, a carousel had a run time of 2:30 and a setup time of 6:18. A car ride had a run time of 2:20 and setup time varied from 3:40 to 4:15. In these examples, the carousel process waste was 72% and the car ride process waste was from 61% to 65%! These were for rides for which there were constant queues! If the setup time could be halved, the capacity of these rides could double! In effect, queues could be all but eliminated!
Some answers to this have already been discussed: elimination of wristband issues, hand stamps, and height measurements, and introduction of SMED. There are also other related issues, such as the operators checking how many people can still fit in the ride. By the way, they regularly miscalculate, i.e. let more people initially enter than fit the ride and then some need to leave, and on some rides they have to walk around the ride to see if there is any more room (this could be solved by cameras or mirrors). In fact, even the decision whether the ride should be filled to capacity or not is one that could be answered by gathering more data: if the setup times can be reduced enough by operating the rides at less than full attendance, such operation could in fact improve the overall capacity.
We ate at a pizza buffet, and it too was plagued with queues. The operations there did not seem too efficient, either.
There were sort of two queues for the pizzas, but they were on one long table and not marked at all. Therefore, it occasionally happened that one queue was very long while the other one was empty, as people thought that the table only had one queue. Some were surprised when the same pizza flavors showed up the second time..
The kitchen was behind that pizza table, but one member of the staff was on the same side as the customers, darting between them and ordering new pizzas from the kitchen and setting them up. With small rearrangements, it should be possible to make the table accessible directly from the kitchen side as well, so that no one would need to dart between the customers all the time.
I did not get a good view of the order management system, so the issues they had could have been capacity issues. Nonetheless, for a buffet, a pull scheduling system would be appropriate. I won’t go into details though, because I could not get a good enough understanding of the current process.
Continuous improvement in an amusement park
Based on all I saw, I do not believe Särkänniemi to be a continuously improving organization. A simple comparison to Legoland, which is located in a much more competed area, shows striking gaps in performance.
However, the main point is not to copy whatever Legoland has done. After all, we already noticed that there are important differences that may make the use of wristbands a good idea for Särkänniemi, even though they are not needed at Legoland.
The main point is to build a continuously improving organization, in which operations are always improved and the current state is never enough. Särkänniemi, for example, has been criticized a fair bit on its price raises for this season. If the organization was able to improve more, they could realize the same profit level with lower prices. Or, well, they could still raise prices and gain even more profits.
The major difficulty an amusement park faces is that it relies a lot on temporary employees, because much of it is open only in summertime. It is more difficult to do continuous improvement if your employees are changing all the time. I will explore that subject some other time, but for now I do have one more thought to share: Imagine if all those students who are working in summer jobs at Särkänniemi learned that continuous improvement is a vital part of work. Imagine the benefits to their future employers, and to their own careers. Whoever could teach them that lesson would have a lot to be proud of.
Photo: Rollercoaster Tornado at Särkänniemi by David Pursehouse @ Flickr