Why BallBots are the future for Indoor Robotics

Robots that are to be used in indoor environments such as offices, houses, hospitals, etc pose many challenging factors that need to be taken into consideration. 

A few examples would be cost, maneuverability, speed, size, traction, robustness, and how intimidating it looks.

Based on those factors, I have decided that the route going forward for indoor robotics is to use develop and use BallBots.

Why? Read on and find out

What Even are BallBots? Sounds like a childs toy

Well, a ballbot is a robot that uses a sphere (ball) to move around and balance on. An example of a ball bot could be the BB-8 however the way it moves around is slightly different to how this planet makes BallBots.

A quick demonstration of a BallBot in action :

Later on in the post, I will discuss how it operates, its pros and cons.

What Type of robot is generally best suited for indoor environments

Before I even get into discussing why BallBots is the way forward, we need to decide which type of robot generally is better suited in our everyday lives.

Bipedal Walking robots are currently far too expensive and have terrible self-stabilizing properties. While they can climb over challenging terrain much as we can, the cost, fragility and overall unpredictability are no good for general peoples use.

So far we have quite a few walking robots but of those, the ones that show the most promise are Cassie and Atlas

Cassie
Atlas

However, with Cassie being sold for just under $300,000 and Atlas being well over $1 million, it seems we just aren’t ready for bipedal robots.

However, there are a few other options such as 2 wheeled balancing robots. Now unlike bipedal robots, self-balancing ones are inherently unstable. So they constantly need to balance themself with complex PID loops and high accuracy IMU sensors.

But compared to bipedal robots, 2 wheel self-balancing bots cost less to nothing. In fact, their whole system can be made up of just 2 robot actuators for the ankle motors. 

Top examples of 2 wheel bots like this are Ascento and Handle. 

Ascento
Handle

Now while wheeled robots clearly have fantastic speed and agility, there are 2 major issues.

Those are: not enough instant manoeuvrability to self stabilise in all directions and it looks slightly intimidating.

You see, when robotics first does start coming out in hospitals and offices, there will be people that want to push around the robot just for fun. The robot has no life and to a random person, it may even just be considered a toy at first.

This means if a robot does not have the ability to self stabilize in all directions as fast as it possibly can then it will have a hard time proving robotics is worth the price. 

2 wheeled robots are unable to stabilize if pushes parallel to their rotational axis

In commercial/industrial robotics, it’s all about how much it can do for the price however when it comes to robotics that is used in our everyday life, factors that may seem meaningless need to be taken into account.

The long going stigma that robots will go and kill you and chase after you were all created by Holly Wood and unfortunately is still present. 

If a robot doesn’t look gentle and friendly then people will shy away from the idea of using it in their lives. This is what bipedal robots and 2 wheeled also robots fall short in.

Another option would just be a generic boxed robot with wheels however that does not have the ability to stabilise itself

Amazons delivery robot. Cannot stabilize itself is pushed over

Now what is left, is BallBots. So why are they better Overall?

Ok, So why BallBots?

Example of a ballbot to get an idea

So far, we have decided that bipedal robots and 2 wheeled robots have too many flaws to be used in our everyday environments.

So here is why BallBots are the best:

  1. Very high maneuverability due to 2 axis inverse pendulum stabilizing
  2. Does not look intimidating
  3. Can maintain orientation while moving around
  4. Low cost

Maneuverability 

Because ballbots move around on a ball, it means that when pushes in any direction, they can instantly stabilise using fast Inverted Pendulum mechanics. 

The Inverted pendulum style of stabilising allows robots to quickly stabilise by placing their body at an angle so that they can shift their centre of mass in opposition to the direction they were pushed.

Here is a 2 wheel balancing robot that also uses an inverted pendulum. It tilts to shift its centre of mass so that it does not fall over

So a ballbot does this in 2 axis compared to the 2 wheel self-balancing robots 1 axis.

 

Does not look intimidating

The ball really just looks harmless. While this doesn’t seem like an important reason, it really is when it comes to having the long waited robotic revolution. A product is nothing if there are no customers.

 

Can mantain orientation

For robots to really be worth the price, they need to do a job as good or better than humans. Ballbots can maintain their orientation while moving around which can be useful when having them deliver files or meals in tight spaces

On the other hand, 2 wheeled robots would have to rotate which takes more time and requires more space

Low cost

Ballbots really just use 3 Omni wheels rolling on a ball to move around. So while it is slightly more expensive than 2 wheeled robots, the price is nowhere the same as bipedal robots.

 

How does it work?

Inside a ballbot for visualization

The robot moves around by pushing the ball using Omni wheel motors. This in turn pushes the robot. The direction that the wheel moves in is decided by the speed and direction of all 3 Omni wheel motors turning. 

Its self stabilising comes from an IMU (inertial measurement unit) and PID loops. The IMU tells the robotics angle and it uses PID loops to gently self stabilise itself. 

 

What are the disadvantages of this?

Unfortunately nothing is perfect and so while ballbots are the best solution, they do have some problems

These issues are:

  1. The ball that it rolls on is detachable
  2. The estimated robot position may be completely different from the actual position due to imperfect wheel traction
  3. Complex coding to move the robot around due to Omni wheels
  4. Need to fine-tune PID loops

 

Overall, I think the benefits of using ball bots outweigh the disadvantages and so Ball Bots are the best choice for indoor robotics. Fantastic stabilisation, fast manoeuvrability and it looks appealing.

 

That’s all from me today. Follow me on my socials to hear more!

linktr.ee/HassanPortal

2 comments

  1. So like those toys that you can knock but they always stand back up? And what about not making the bot very tall but making it bottom heavy (like those buildings which are designed to be less likely to topple in an earthquake) with 3/4 wheels instead of just two. Interesting to read about the ball’s stability though

    1. Yes I guess It is sort of like a fidget stick. Also shifting the centre of the robot of mass downwards would make it more stable but it would still have to constantly be self-stabilizing. If you want a more stable design, you could always go with standard boxed robots that have a wide base however my issue with that is there isn’t a way for it to regain its normal position if it were to fall over.

Leave a comment

Your email address will not be published. Required fields are marked *