15 March 2016
13 March 2016
Seek and Geek #6: Best of ABC's Battlebots 2015
The most recent revival of the Battlebots TV show on ABC was pretty exciting to watch, and this year there are a few teams affiliated with MIT that will be competing in the 2016 season. I'm happy to say I'm helping out a bit with one of those teams, a bot called The Dentist. While most of the design has been done already by the core team members, I've been helping out with some machining and general integration.
I'd like to take a look at some of the most successful and worst robots from last year's season, analyze what worked, and what didn't. While it's too late for me to make use of these lessons in designing The Dentist (I was roped in after most of the design was done), I think this will be useful for engineering in general, if not for a future battlebot.
Successful Bot: Tombstone (or Last Rites)
Tombstone's Strategy: Have the best weapon, maximize its kinetic energy.
Defending against Tombstone: be fast to avoid impact, have armor designed to take a direct impact, have enough traction to stay put when hit
Tombstone (and its previous iteration, Last Rites) is ridiculously powerful. It chews through anything by having as much of the total 250lbs allowed in the bot as possible in the weapon. The more mass the weapon, and the faster it spins, the more kinetic energy it has. The more energy you have, the harder you can hit, and you can brute force your way through any competitor.
The issue is, however, Tombstone needs to be able to handle its own impacts. If it hits another robot and sends it flying, it may be able to handle the reaction impulse, because the rotor will not stop spinning. In the slow motion video above, you can see it chew through furniture with minimal reaction force (the furniture just melts away).
If Tombstone hits a wall, however, the rotor has nowhere to go because it cannot push the wall away, and so the robot takes the brunt of the impulse and sends itself flying. The bearings and structure need to be able to deal with this, as well as the electronics within. When Tombstone is successful, however, it is terrifying to watch, like when it went up against Counter-Revolution.
Speaking of which...
Unsuccessful Bot: Counter-Revolution
The gist: weak weapons, armor cannot handle impacts, weak drivetrain.
This sad robot looked pretty cool at first. It had two vertical spinning weapons, but they never even got to do any damage before Tombstone came in and made quick work of the thin aluminum armor. The drivetrain was also not well-designed, as it is unable to provide any traction if the bit is lifted up at all, and is too weak to be able to maneuver and dodge Tombstone.
Successful Bot: Bite Force
The gist: RELIABLE. Strong armor, strong drivetrain, re-configurable weapons.
Bite Force ended up winning the overall competition! It has a beefy frame of either steel or aluminum, which takes a beating and never lets up. I don't think it ever took any actual damage throughout the entire competition. In this video it's up against MIT's Overhaul, and while they have similar designs, reliability is KEY. Bite Force can lift other bots without ever losing traction due to reliable design. Bite Force can take a monstrous beating without ever having a part break because the armor is thick as hell and is well-designed. Bite Force's lifter can be strategically reconfigured depending on the immediate competition to either be a lifter primarily, or a wall of steel with very few appendages.
Here is the championship fight, and Bite Force was a match for Tombstone because it was the worst thing Tombstone could possibly come up against: a wall of steel. Tombstone is normally relentless because it can chew through most armor it comes up against, but Bite Force has been designed to take a worst-case direct impact from Tombstone without taking any permanent (yield) damage.
Neutral Bot: Witch Doctor
The good: Vertical spinner with high kinetic energy: can throw around Tombstone. Nice front armor.
The bad: Unprotected wheels, no means of self-righting when flipped upside-down.
Witch Doctor is pretty cool, because it gave Tombstone a run for its money. It actually broke Tombstone's main weapon! Its compact vertical spinner was effective in throwing around Tombstone like a toy, and the surrounding armor held up quite well against Tombstone. Its cantilevered wheels were right out in the open and unprotected, however, and it was unable to self-right when flipped over, leading to its loss against Tombstone.
06 March 2016
Seek And Geek #5: Hydraulic Cherry Picker
The other day I saw a Hydaulic Cherry Picker like the one above parked between Kresge and McCormick which had been used to put the "MIT 100 Years In Cambridge Celebration" posters up on Building 7.
The whole thing has a telescopic arm made of square extrusion. I couldn't get a good peek inside of the mechanism it uses to actuate the telescoping mechanism, but considering everything else on this machine is hydraulic (which is good because there is only one power source that is distributed across the whole machine) I bet this is probably actuated hydraulically, with the piston motion amplified in some way through gearing or a linkage.
Speaking of linkages, the whole thing has a "wrist" that's on a parallelogram 4-bar linkage, so it keeps the basket holding any operators parallel to the ground.
A workspace analysis found online shows off just how much reach this thing has. The wrist allows fine positioning once the big arm has been extended.
The big arm is actuated by a massive hydraulic piston in a 3-bar linkage. The forces it has to exert to hold the arm in place change given the angle of the arm, and the horizontal configuration requires it to bear the highest load. Luckily, static force holding is something inherently free in hydraulic systems!
You can see the cable/hose carrier, bringing both electronic cables and hydraulic hoses up the arm to the the pistons located near the basket.
The wheels are hydraulic! Both drive and steering are achieved through hydraulic actuators. This thing is really slow when driving, no more than 15MPH, but at least no additional forms of power are needed. The same hydraulic power unit is used to move the arm, the basket, steer, and drive the wheels.
There are two pistons that orient the basket! One is coupled to the 4-bar parallelogram linkage, and the other will simply change the angle of the entire basket with respect to the arm. This is probably coupled to the motion of the arm's beefy main piston near its base, so the basket is always oriented horizontally.
There also is a left-right rotational joint at the basket's wrist, also hydraulically actuated.