Last year, we visited several customers to do deep field research, including Huaxin Cement, Zoomlion, and South Cement. We interviewed the customers and had design thinking workshop with them, get better understand of their business process and figure out some pain points and needs in the flow.

The users
Our users are mixer truck drivers. Their main goal is to finish all the transportation tasks on time and safely every day, and take care of the truck including maintenance and energy saving.
When the dispatcher finish preparing the schedule, the tasks will be numbered and displayed on a large screen, meanwhile integrated with production system. The drivers need to take turn to execute the transportation after each batch of production. Sometimes the drivers have to wait for a very long time for the next round task because the assignment seems extremely flexible and even somehow out-of-order. When a driver receieves a task, he has to drive to the concrete discharge hole and get the invoice which is delivered by a very traditional way - passed through a pipeline from upstairs to downstairs holded on a clip tied to a rope. After arriving at the construction cite, he may also have to wait for the trucks before head to finish their unloading firstly, this always takes a long time. 

Besides finishing the tasks, the other thing the driver must take care is the mixer truck, including its health and oil consumption. Nowadays, maintenance starts with something wrong and brings much down time which makes the resource reduced. Also, high oil consumption produces high cost, but the reason is not clear and transparent enough. But if we use the technology of IoT, we can collect real time data from the trucks by sensors and tell the driver when to maintain the vehicle, for what reason they generate high oil consumption, these can greatly help the drivers do a better work.

The main problem for the drivers is fragmented information which is non-digitalized, and lack of real time monitor information of the truck. It brings lower work efficiency and higher mistake rate. For an ideal condition after the dispatcher finish scheduling, the tasks should be sent to each driver's app as a task list. During the transportation, since the before head truck unloading information is not able to be collected into the system, they may need to communicate with the front truck driver in the queue to know the real time condition, in order to adjust driving speed and reduce waiting time. Also, they may get reminder when the truck needs maintenance, and see the reason of high oil consumption which may due to their own bad driving behavior.

After the first round wireframe prototype, the usability testing plan was made before iteration. Since we have very limited resource and budget for formal usability test, we prepared a DIY one. What does DIY usability test mean? It means informal, small sample, and simple. Go out to find either target users or non target users to go through the and try the app prototype, most of the serious usability issues can be figured out and improved.

The testing task list including the 7 parts is arranged from easy to hard, which is easier for the interviewees to get started. Totally we did two-round testing and iteration. The problems, suggestions mentioned by interviewees, modification plan, problem category, corresponding usability principle have been recorded for analysis.

According to the feedback, the design will be changed correspondingly, but we need priority, not change for all the points. Therefore, we put the feedbacks into a matrix as below:

Research, analyze and define, ideate, prototype, iterate and improve, these compose a complete design process, and usability test should be throughout the whole process as a very effective way to develop a contextual design in a very convincing way. It's easy to get started but difficult to do well since you may encounter all kinds of unexpected feedback through the testing. 

Some impressions:

1. When facing conflictive feedbacks, we should have our own judgement by analyzing the user's background, character, mood and situation, then make a decision whether to take the feedback into account. 

2. Remember to start the testing as early as possible in each step, if the fail is inevitable, fail early.

3. Keep an eye on the user scenario and deeply think about it. How many situations does the user have? At what condition dose he open the app? Why should he see this page? What is mostly possible for him to use this function? Keep thinking about these questions and adjust the design again and again.

4. When it seems impossible to find real target users to do the test, find someone else who has similar features (for expamle: who drives) and has no knowledge of your project, you can also get many valuable feedbacks from them because typical usability issues can be figured out by anyone.

 

 

It's good to do usability test for improving design, we can get more insight and sense of how to avoid typical usability issues and how to make it better as time going on.