What was this beautiful thing you created?
Kater Technologies Inc was a Vancouver-based Mobility-as-a-Service (MaaS) technology company that launched an integrated, intermodal transportation network designed to combine different forms of transportation into a single service, making taxis, ride-hailing, ferries, scooters, buses, and even planes all integrated, all working together in a single app.
The platform allowed users to book a trip filtered by cost, duration, arrival, or departure time and then be issued a boarding pass that handles all payments to the various transportation providers. The net result is cost savings, time savings, and ease of use. With our top priorities being safety and customer service, the Kater app was designed to create a superior user experience and improve transportation for British Columbians.
Product Design Lead, Researcher, & Communicator
- Product Design LeadFrom the customer’s goals and motivations, I start out with the user story and map out the product scope for our specific target users which I then carry out with the product team that I manage. We then go through the product design cycle of wire-framing, prototyping, user testing with customers and drivers, iterations, and getting the engineering team’s feedback. I also audited and signed off on completed features when confident it is ready and then I would monitor its success through data and/or community and customer support channels.
- ResearcherI have traveled to Seattle, LA, and Portland to do competitive analysis research on Uber and Lyft and presented the findings to key stakeholders to direct our strategic development of the company. One example included researching a new method to help users and airports direct the flow of traffic at pickup and drop off points. Upon developing a prototype to solve this issue we gave a presentation to our executives which was then presented to Vancouver YVR’s Airport Authority.
- Communicator for Multiple DepartmentsI had worked on a daily basis with the CTO, senior mobile leads, backend leads, the UX/UI lead, Product Designers, the Director of Engineering, the VP of Operations, and the Director of Marketing to make sure we are all on the same page and able to execute the vision to everyone. In one year a lone, I’ve worked on 60+ features for four mobile apps (Customer/Driver for both iOS and AND) and a Control Centre/Website, basically being “all over it” in terms of the status of every piece.
Mapping is very very hard
With Kater being a ride-hail company we relied heavily on proper mapping, routing, and navigation however our decision to use Mapbox and our ambition to get the application out the door had many hard lessons to learn from. Here are a few:
1. Displaying Softmeter Costs
One of the main reasons we chose Mapbox over Google Maps is because as part of the government’s transportation policies if we were to use taxi licenses is that we needed to display the cost to the customer at all times. With this, we had problems with the soft meter displaying the incorrect cost due to low network, no data, having a VPN, not enough traffic data in Canada, and mobile bugs that prevented the proper uploading of the coordinates to the server. Our workaround for this was refactoring the mobile apps but also putting a guard on the mobile side if the network is lost, it stopped calculating the coordinates until it is back online again and would play catchup with the roads crossed.
2. Inaccurate Point of Interests
Before launch, we had a really weird issue that brought the driver and customer to the center of the airport tarmac as opposed to the entrance/exit ways. I was put in charge to communicate with Mapbox and figure out where the lat and long points are supposed to actually go. Working with the head of the mapping division we discovered because we are using Google’s Search API to give us the search results and Mapbox’s navigation SDK it created a conflict where Mapbox’s navigation would take the user to the middle of the Point of Interest (POI) drop because that is where is the Google’s pin-drop is. To work around this, we made “Kater Maps” a tool that allows the admin to populate what POI name we want to call it and more accurate lat and long from looking at Google Map’s coordinates. The end result was we were able to populate both arrival and departures for domestic & international entrances and create even more if there are other entrance ways. This feature would also be a bonus in the future if we wanted to work with businesses that have multiple entranceways for E) Fairmont Pacific Rim North Entrance.
3. Inaccurate traffic Data, ETAs, and routing
Mapbox is an American company that relies heavily on open-source map data which has a major lack in Canada. So when we rely on Mapbox to give us accurate traffic data, correct routes and ETA’s of when the driver will be arriving and dropping them off it came out inaccurate. There were many times I had to be in contact with Mapbox when we were in the field testing and the driver was taken to a dead-end, their support would tell us to go to open street maps and update it on that end. For routing the maps division tried using Graph Hopper which is a fast Directions API with worldwide data from OpenStreetMap and route optimization. It helped with routing a bit but with the lack of traffic data in Canada, it still was problematic.
4. Invoices being inaccurate due to driver’s VPN
Upon release, we were getting customers calling in that they were charged almost a million dollars for a 20 min trip. The culprit was that one of the drivers was using a VPN that was located in Asia so when they were making a trip, the mobile’s GPS device thought the driver was in Asia thus causing the CSV to generate an in-accurate cost to the actual trip. The workaround was to create a guard that would prevent a CSV from uploading if the vehicle was going faster than 150km/h.
5. Curbside pickup
Another major issue we had was customers not being picked up the curbside or outside the entrance of their building especially with downtown offices that has multiple exits and entranceways. To solve this, the maps division lead, product designer and I worked on “Forced pin” which allowed the user to select an option if there is more than one pick up location. Or the user would be able to drag the pin as it it is floating and it would snap to a building entrance that is populated on the map to showcase the multiple different entrances.
6. Google being the golden standard
We had a lot of pushback from drivers and the operations team for why we don’t use Google for our mapping provider. Two things that were the main reason was because of significant costs and the ability to show the soft meter on the screen. Eventually, when ride-hailing came to BC we didn’t need to show those costs anymore. So the design team and I came up with some solutions on how we can allow the driver to navigate using Google Maps but also have the Kater App running in the background and easily be able to click the widget on the side to be able to get back to onto the screen and vice versa all while doing this while driving. During the development, we needed to test a lot with drivers and think about the road conditions they are in if they are left or right-handed, and resolution size for drivers who need the map to be bigger as well as typography.