Jacky’s Business Solutions key to deploying robot customer assistant
As COO for Jacky’s Business Solutions, Ashish Panjabi, plays a decisive part in taking robots from SoftBank Robotics out into the Middle East market. Appointed as a value added reseller earlier this year, Jacky’s helps end-users effectively deploy SoftBank’s customer-advising humanoid robot, Pepper, in their businesses.
Can you explain a bit about that process of taking the robots to the end-users?
We have two channels – one is a direct channel, the other is working through reseller partners. Ideally, we’d like to work as much as we can through the partners because a lot of the work is the last mile in terms of software development, software tweaking and improvements are coming in over the period.
Again we compare a lot of it (progress) with the mobile app. You started off with a mobile app 10 years ago and it was based on a version of the web page and that mobile app kept evolving over the period because once you had the app you said ‘wait a minute can the app do this, or could I be doing this off the mobile?’
And eventually one by one you start to see more and more features as that mobile app goes from version one to one point one and one point two until it hits version two and version three. Then it’s completely different to what it was 10 years ago. So we see a lot of this with the robots as well as people taking it, deploying it and thinking ‘could I be doing this on it.’
For instance, in a retail store you could use the robot as part of your catalogue. It could be something that links to a loyalty system that uses a camera so it could recognise you when you walk in the store.
This could alert the store staff that you were in the store otherwise they only find out after you paid. They could make recommendations based on your history. It also could be used to finalise a transaction.
How are governments using the robots?
One of the best use cases we’ve got is DEWA, the electricity and water authority. They started with five Pepper robots and initially they took a third-party developer. They put it in the branches and they were getting feedback.
They also had to think what more could Pepper be doing. It got to a point now where they looked at it and said ‘can we start developing in-house, because you’ve got such a long laundry list of things that we think the robot could be doing here’. They said ‘it’s more efficient if we do it in-house because we understand what needs to be done.’
So we actually had them trained. We had the programmers from France come in, we held training sessions for DEWA. They’re now fully self-sufficient and they can do all the development work in-house.
They do all of their systems, they work with all their partners and they are buying more robots as well. Their focus is they are investing in customer happiness centres or branches and they don’t want any humans in them.
So it’s a matter of changing habits because as long as there’s a human there you’re going to go to a branch and say can you do this for me or can I pay you? Or could you disconnect my electricity? What they want to do is use the technology that’s in the stores so eventually you start doing it off your mobile app and you don’t come in anymore. They need to get you in the habit for that.
What form does your training take?
For the various things on the software side, most of these robots work on Python. So it’s programming the Python language so you understand the way the robots work because it’s not only the software processors or the transaction processors that are in there. The robot could speak to you in a monotone voice but if you want it to move a bit and you want some emphasis than you need to program all of that.
You’ve got 150 sensors in the robot. It’s also a question of being able to enable and use the right sensor at the right time.
For example, you have something called a mood mirror. Basically this is a sentiment analysis. So you can be talking to the robot and it could realise whether you are happy or not happy. Because it starts to recognise certain patterns and if you start linking this to a Machine Learning system, or an AI system, this could actually realise if you’re in a bad mood. Then it could crack a joke, just to lighten things up or it could do a little dance or something.
When you take this into the market and, for instance, you’re trying to sell the robot to a bank how do you do that? Do you just explain how the robot can interact with the customers?
The most basic things is it needs to solve a problem. We have to understand what the problem is. What we’re starting to see a lot more with our police enterprise clients is they are going through design thinking workshops in any case and out of these they start identifying a lot of the pain points. This identifies the different things that could be done better.
When someone walks into a bank and they see the robot what sort of reactions are people giving to it?
I think it’s sort of like a cross section that you have in the country where you have people of all levels and all types. Some people love it, some people hate it, you’re going to have that. But I think the one thing is we are in a fairly progressive society, Dubai has been a society of influencers and thought leaders and I think the rest of the region looks at it to see what has been happening.
I think from a bank perspective, when one bank does it the others start looking and saying ‘what could we be doing and how can we get the edge over the other guy?’