Case Study: Stop! Wait! I Am Pulling Down a Menu!
San Jose, California, is considered one of the safest large cities in the United States. The city’s 1,000 police officers serve 925,000 residents, making it the smallest officer-to-resident ratio in the country. From 1990 to 2004, the city’s police department used a text-based mobile dispatch system. That system had been customized by its designer to meet the needs and preferences of the city’s officers. Although there was some initial hesitation by officers to use the system, they eventually embraced it. After more than a decade of reliable service, police and city officials decided to replace the system with new Windows-based touch-screen software. A new touch-screen computer was to be installed in every patrol car. It was designed to receive orders, send messages, write reports, receive maps of the city, and use GPS to let officers know where they are located and where other patrol cars are. San Jose government paid Intergraph, the company that developed the software, $4.7 million for the software, which was supposed to serve both the police and fire departments. However, the effort was plagued with problems from the start. Even before the new system was installed, there were already grumblings at the department. Officers claimed nobody had ever sought their input about the design of the user interface. When they started using the new system, they were disappointed. Tension had built up, but this was not the main concern of the San Jose Police Officers ssociation (SJPOA). The organization’s leaders were not so much offended because they had not been asked about the system before it was developed. They were more concerned about the results of that failure to ask for their members’ feedback. They were frustrated with the lack of training and error-infested software. There are always complaints when people have to adapt to new technology, but when their lives and those of the public depend directly on the software’s performance, the stakes are much higher. Since its June 2004 operational debut, the system has had numerous major problems. Of greatest worry is the increased difficulty in issuing the Code 99 command, the emergency contact when an officer is in danger and needs immediate help. Initially, officers had to strike one key to issue Code 99, but that resulted in too many false alarms. As a result, code entry for emergencies now requires a two-keystroke combination. Officers complain about having to find the right combination of touch-pad keys on a 12-inch screen while they are under fire or in hot pursuit of a suspect. One officer even crashed his squad car into a parked vehicle because he was so distracted by the information he had to enter using the touch-screen. Another problem was that with the new software it took patrol officers longer to find out whether a person they have stopped has a violent criminal record, which is vital information in a job that requires split-second decisions of life or death. The police officers complained that they were not given sufficient training. However, the problems with the system had nothing to do with how the police officers used it; the software simply did not work. Two days after the system went live, it crashed. For the next few days, it was almost completely inaccessible. Its designers acknowledge that this was not a good way to build confidence with the officers. Yet, even after the system was modified to fix these problems, several more errors were discovered by the president of a userinterface design consulting firm that was hired by the SJPOA to review the software. The mapping and GPS location tracking were supposed to be a great helper. Yet, the system’s map information had some significant inaccuracies. Additionally, unneeded information took up screen space and display fonts were hard to read. Even a simple task such as checking a driver’s license plate was difficult to perform after the system had already been treated for bugs. Every new technology has a learning curve that can last weeks or months until users feel sufficiently comfortable with it, but with this software the difficulties were not only a matter of a learning curve. Even tolerant and receptive officers have faced obstacles in trying to adapt. Intergraph’s specialists spent weeks in San Jose to fix bugs and streamline procedures for the most basic patrol tasks, like the license plate verification. Officers complain about receiving only 3 hours of training on software that is supposed to ensure their safety. In response, the department has offered more training sessions. The software runs on the Windows operating system, a fact that complicated matters for many of the police officers. Older officers are not comfortable with pull-down menus and other features of the interface. As a result, they have been more resistant to the new software than their younger, more computer-literate colleagues. Observing police work, the consultants brought in by the SJPOA noted that choosing a Windows GUI with complex menu hierarchies does not make sense for anyone who has to use the system while driving a car. In addition, officers were trained on desktop computers with track pads on keyboards instead of touchscreens they actually have to use in the squad cars. Dispatchers, too, have expressed dissatisfaction with the Intergraph system, especially because of risky delays in task execution. With the new software, officers have to wait longer to access information about any previous arrests for a detained suspect. Dispatchers also note the same concern expressed by their comrades on patrol: the new software cannot perform multiple tasks simultaneously. Like the officers, the dispatchers feel they should have been consulted about the software during the interface design stage. San Jose’s police chief admits that in hindsight, incorporating more end user input during the planning phase would have eased the introduction and implementation of the new system. The Chicago Police Department had a similarly painful experience with a major dispatch system overhaul in 1999. Just as in San Jose’s case, patrolling police officers were not asked for input before the software was developed, and the results left bad feelings across the department. Chicago eventually replaced that software with a newer system. This time, patrol officers were consulted, and their suggestions were considered before the programmers developed the applications.
Unfortunately, San Jose’s police department did not learn the lesson from the windy city’s experience. Police departments in two Canadian cities, Calgary and Winnipeg, had similar disappointing experiences with the Intergraph system. Officials in other cities also have been frustrated, and some planned to scrap the system. Perhaps San Jose might not have to replace the Intergraph software after all. The San Diego Sheriff’s Department has used Intergraph’s touch-screen software for 6 years with eventual success. Initially, there were bugs similar to those experienced in San Jose, but Intergraph eventually fixed them. Also, San Diego officials conducted basic Windows training sessions for their sheriff’s deputies, because some of these people had no previous computer experience whatsoever. There was some resistance to the new software in that department, too. But fixing the bugs and providing good training did the trick, and the deputies adapted. San Jose city officials hope for a similar happy end. However, unlike San Jose, San Diego squad cars do not have touch-screen devices.
Source: Hafner, K., “Wanted by the Police: A Good Interface,”
New York Times, Technology Section (www.nytimes.com), November 11, 2004; Zapler, M., “New S.J. Dispatch System Flawed,”
Mercury News (www.mercurynews.com), September 22, 2004.
Questions to this case:
1. Are the problems encountered by the police officers due to hardware or software?
2. Whom do you think is at fault for the unsuccessful implementation of the new software? Why?
3. People, especially the “technologically challenged,” are often not receptive of new technologies. Was this a major issue in this case?
4. If you were the CEO of Intergraph before it assumed the project for San Jose, what would you do differently?