Philip Mason talks to Motorola Solutions’ CTO Paul Steinberg about the product development process, the company’s relationship with emergency services, and what public safety might look like in 20 years’ time
CCT: Has there been an increasingly collaborative element when it comes to the development of new solutions for the public safety sector?
PS: Nowadays emergency services users need to do more with less and technology can help them be more efficient and effective, but it can only be a powerful tool if it is intuitive and embedded into processes and systems. Experience has taught us that user research is incredibly important – in addition to the classic market research approach.
A good example would be the work we’re doing with a customer and Neurala, a Boston-based AI start-up, to develop “intelligent cameras” for public safety users. The goal is to enable police officers to more efficiently search for objects or persons of interest, such as missing children and suspects. We now have a working prototype of the technology that we are using to gather feedback from customers.
CCT: What form does that user research generally take?
PS: The method we use is embedded ethnographic research, which involves a team of people with a background that is a mix of psychology and anthropology looking at the ‘human’ side of the equation and the factors around that. Our global teams spend time with public safety organisations to closely observe how they work and also conduct in-depth interviews about the processes and tools they use to do their jobs. This research leads to insights about customer “pain points” or jobs to be done. Those insights guide our brainstorming and development of proofs of concept, after which we move into co-creation with our customers. We work closely with the customer to develop the exact solution they need and ensure it is one they can intuitively embed in their processes.
CCT: Why is that kind of information important?
PS: It’s important because it helps us get to the right solution in relation to what the end-user is trying to do. At the same time, it generates perspective from which other inspiration might come. There may be a pain point that the users didn’t even know they had but they’ve shown us as they go about their business. We often say that customers can’t always tell you what they need, but if you watch them the right way, they’ll often show you. Ethnographic in this context essentially means empathy for the human condition. We try to understand people on their own terms.
CCT: How has that worked in practice? Could you give me an example – for instance, in relation to something like Pronto.
PS: Pronto was started before we acquired Airwave, but it’s evolved since then. Initially, the product was highly customised for each client. We’ve since worked to establish common denominators across the whole user base, because we think that works out best for everyone.
Another example relates to the AI work we’re exploring with body-worn cameras, in which we recently completed some customer research on public safety applications for AI-powered object recognition, which uses AI to search for items matching a given description (such as an adult male wearing a blue shirt). Our team spent time with five police agencies of different sizes and jurisdictions across North America. Some of the research revealed that searching for a person of interest during patrol work is a process with usually low return on investment, and that “being on the lookout” for several people brings a significant cognitive burden – having to store many descriptions in one’s head at the same time. These are the kind of use-cases where AI can help a person do their job more efficiently.
CCT: What are the main concerns of the emergency services at the moment, from what you can gather?
PS: Safety in the first instance – how they can keep their people from harm – followed by concerns around efficiency, as well as doing more with less.
Citizen engagement is of increasing importance to emergency services to improve transparency and interact with the public. Related solutions can help public safety organisations to connect with the public through web and mobile apps, crowdsource intelligence related to incidents which ultimately can help to increase safety in cities and communities.
The Pronto mobile application suite mentioned before is a great example of technology that improves efficiency. Some of our customers reported being able to cut costs by up to £7m per year, by replacing paper-based activities with intuitive digital forms on mobile devices.
CCT: How important is predictive policing going to be going forward?
PS: We believe that it will be extremely important. Police officers tell me how valuable it is to have an officer visible on-scene, because it offers such a level of reassurance. Predictive policing essentially comes down to getting people in the right place at the right time.
We’re currently involved in Detroit with an initiative called Project Green Light, which uses real-time HD video surveillance in high-crime areas that live-stream into the city’s police department. That includes places like gas stations, convenience stores and so on. Analysts monitor the footage around the clock, enabling them to quickly deploy officers where they’re needed most and understand exactly what’s going on at a given time. The cameras also serve as a deterrent, leading would-be criminals to avoid locations where they know they’ll be on camera. There’s several-hundred of these deployed across Detroit, and we’ve seen a significantly reduced level of crime in those locations.
CCT: How are more complex data sources likely to be analysed, and how will that feed into crime prevention?
PS: It’s ultimately a matter of cross-referencing different information sources, either using human beings – as with Project Green Light – or AI. When it comes to AI, let’s say that the system detected a gunshot being fired somewhere in the city, which it then places on a map and triangulates. AI can then relate any calls made around that time to the incident in question, examining the first words of the conversation and so on. That’s a very simple example of how AI can be used to do a reduction on data that is coming in.
CCT: There’s a sense that emergency services have been quite reticent to share information with each other. What are the cultural barriers that must be overcome if this kind of technology is going to be exploited to its best effect?
PS: It’s true that silo working exists, although a lot of that is just how systems have grown up and been managed over time.
In terms of the culture of organisations, we are seeing an increase in understanding of the power of information sharing. That could be between emergency services, but it could also bring in information from the broader social care agencies as well, which often have important data that can indicate the likelihood of an individual’s involvement with criminal activity.
CCT: What’s the public safety landscape going to look like in 20 years as the technology continues to develop?
PS: To me, the biggest limit we have today is that we can only consume a finite amount of information, because we’re limited by what we can hear and see. At the same time, we can still think incredibly quickly, so we could see a situation where we somehow move information between us biologically, crossing the human neural interface barrier. I have to think that will happen.
In the realm of public safety in particular, the use of robotics is likely to become increasingly significant. Our users go where it’s hard, expensive or unsafe to get to, so the business case is already there. It won’t exactly be Robocop, but I saw a use-case recently in Singapore with exoskeletons that firefighters could use to help them carry more equipment.
CCT: How do you avoid coming up with solutions to problems that don’t exist?
PS: It comes down to collaboration, planning and working out a genuine business case based on need.
It’s always going to be a journey.
Author: Philip Mason