MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati 2/7: Knowledge-based Planning
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
2/7: Knowledge-based Planning
2
Life must be lived forward, but can only be understood backward.--Kierkegaard
Lp relaxation
3
Should we bother with Proof-based encodings?
• Clearly Graphplan-based encodings are better than backward-proof based encodings
– Because Graphplan does Mutex propagation– And mutex propagation is harder to do on direct
SAT• And yet, it is useful to know how to set up encodings directly—since if we
don’t know how to do directed consistency enforcement, then we at least know how to solve the problem
– Particularly useful for non-classical problems, where we haven’t yet found really effective mutex propagation procedures.
– If “neural networks are often the second best way of solving any problem”, you can say that direct SAT(or other) encodings are the second best way of solving planning problems
• Which is better than “no way at all”
Connections to Refinement Planning Framework
5
Disjunctive Planning• Idea: Consider Partial plans with disjunctive step, ordering, and auxiliary
constraints
• Motivation: Provides a lifted search space, avoids re-generating the same failures multiple times (also, rich connections to combinatorial problems)
– The “solution check” now involves searching through the space of all (exponentially many) minimal candidates of the plan to see if any of them is a solution
• Issues:
– Refining disjunctive plans• Graphplan (Blum & Furst, 95)
– Solution extraction in disjunctive plans• Direct combinatorial search• Compilation to CSP/SAT/ILP
6
Disjunctive Representations--Allow disjunctive step, ordering and auxiliary constraints in partial plans
< 1,In(A), > V < 1 ,In(B), >
1: Load(A)0
1: Load(B)0
1: Fly(R)0
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A)0
1: Load(B)0
In(x)@
In(x)@
In(B)
In(A)
Load(A)0
1: or
Load(B)
At(A,E)@1 V At(B,E)@1
7
Refining Disjunctive Plans (1)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A)0
1: Load(B)0
1: Fly(R)0
1: Load(A)0
1: Load(B)0
1: Fly(R)0
2: Load(B)
2: Load(A)
1: Load(A)0
1: Load(B)0
2: Unload(A,E)
2: Unload(B,E)
1: Load(A)0
1: Load(B)0
2: Fly(R)
2: Fly(R)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A)
2 : Load(B)
3 : Fly(R)or
or
or
4 : Unload(A,E)or
5 : Unload(B,E)
+ Maximum pruning power- Exponential refinement cost
INFEASIBLE
Indirect unioned approach
8
Refining Disjunctive plans (2)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A,E)
2 : Load(B,E)
3 : Fly(R)or
or
or
4 : Unload(A,E)or
5 : Unload(B,E)
7 : Unload(B,M)
9 : load(B,M)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
union of states
In(A)
In(B)
At(R,M)
At(R,E)At(A,E)At(B,E)
At(A,M)At(B,M)
+ Trivial to construct - Loss of pruning power (progressivity) => too many (minimal) candidates
=> Costlier solution extraction
Direct naive approach Put all actions at all levels --Proposition list contains all conditions
USED in S
ATPLA
N
6 : Unload(A,M)
8 : load(A,M)
9
Refining Disjunctive plans (2)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A)
2 : Load(B)
3 : Fly(R)or
or
or
4 : Unload(A,E)or
5 : Unload(B,E)
or
6 : Unload(A,M)or
7 : Unload(B,M)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
union of states
In(A)
In(B)
At(R,M)
At(R,E)At(A,E)At(B,E)
At(A,M)At(B,M)
+ Trivial to construct - Loss of pruning power (progressivity) => too many (minimal) candidates
=> Costlier solution extraction
Direct naive approach Put all actions at all levels --Proposition list contains all conditions
USED in S
ATPLA
N
10
Refining Disjunctive plans (3)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
+ Polynomial refinement time-- Loss of pruning power (progressivity)
Enforce partial 1-consistency Proposition list avoids unsupported conditions
In(A)
In(B)
At(R,M)
At(R,E)At(A,E)At(B,E)
At(A,M)At(B,M)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A,E)
2 : Load(B,E)
3 : Fly(R)or
or
or
4 : Unload(A,E)or
5 : Unload(B,E)
7 : Unload(B,M)
9 : load(B,M)
6 : Unload(A,M)
8 : load(A,M)
11
Refining Disjunctive plans (4)
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
In(A)
In(B)
At(R,M)
At(R,E)
At(A,E)
At(B,E)
+ Polynomial refinement time higher pruning power+ Better balance between refinement cost and solution extraction cost
Enforce (partial) 2-consistency Proposition list maintains interactions between pairs of conditions
1: Load(A)
2 : Load(B)0
3 : Fly(R)
or
or
1: Load(A,E)
2 : Load(B,E)
3 : Fly(R)or
or
or
4 : Unload(A,E)or
5 : Unload(B,E)
7 : Unload(B,M)
9 : load(B,M)
6 : Unload(A,M)
8 : load(A,M)
Can do higher level consistency, but
it is rarely worth the cost
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Digression: Domain-Independent vs. Domain Specific vs. Domain Customized
Domain independent planners only expect as input the description of the actions (in terms of their preconditions and effects), and the description of the goals to be achieved
Domain dependent planners make use of additional knowledge beyond action and goal specification – Domain dependent planners may either be stand alone programs
written specifically for that domain OR domain independent planners customized to a specific domain
– In the case of domain-customized planners, the additional knowledge they exploit can come in many varieties (declarative control rules or procedural directives on which search choices to try and in what order)
– The additional knowledge can either be input manually or in some cases, be learned automatically
Unless noted otherwise, we will be talking about domain-independent planning
Rev
iew
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Improving Performancethrough Customization
Biasing the search with control knowledge acquired from experts– Non-primitive actions and reduction schemas – Automated synthesis of customized planners
» Combine formal theory of refinement planning and domain-specific control knowledge
Use of learning techniques– Search control rule learning– Plan reuse
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
User-Assisted Customization(using domain-specific Knowledge)
Domain independent planners tend to miss the regularities in the domain
Domain specific planners have to be built from scratch for every domain
An “Any-Expertise” Solution: Try adding domain specific control knowledge to the domain-independent planners
ACME
all p
urpos
e
planner
Ronco
Block
s world
Planner Ronco
logist
ics
Planner
Ronco
jobsh
op
Planner
AC-RO
Custom
izab
le
planner
Domain SpecificKnowledge
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Domain Knowledge—Prescriptive or Suggestive?
One big question is whether the additional domain knowledge given to the planner – suggestive? (soft constraint on the search and solution)– OR– Prescriptiive? (hard constraint on the search and solution)
At the outset, it would seem like either way should be fine. However, handling soft-constraints correctly is harder than handling hard-constraints..
So, most knowledge-based planners use the domain knowledge as “prescriptive”
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Task Decomposition (HTN) Planning The OLDEST approach for providing domain-specific
knowledge– Most of the fielded applications use HTN planning
Domain model contains non-primitive actions, and schemas for reducing them
Reduction schemas are given by the designer– Can be seen as encoding user-intent
» Popularity of HTN approaches a testament of ease with which these schemas are available?
Two notions of completeness:– Schema completeness
» (Partial Hierarchicalization)– Planner completeness
Travel(S,D)
GobyBus(S,D) GobyTrain(S,D)
Getin(B,S)
BuyTickt(B)
Getout(B,D)
BuyTickt(T)
Getin(T,S)
Getout(T,D)
Hitchhike(S,D)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Modeling Reduction Schemas
GobyBus(S,D)
t1: Getin(B,S)
t2: BuyTickt(B)
t3: Getout(B,D)
In(B)
Hv-Tkt
Hv-Money At(D)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
(Thoroughly) Mixed Motivations for HTN Planning
To model Partial domain knowledge (when the domainWriter doesn’t haveFull domain model)
To provide a grammarFor desirable solutions
To allow for use of Abstraction/Hierarchy in modeling
Abstraction used mostly for speedup
HTNs provde a naturalSubstrate for this,But there has beenLittle formal use of HTNs this way
Unless we put strong restrictions on HTN schemas it is hard to connect any notion of abstraction—such as Downward refinability-- to HTN schemas. (Since HTNs can allow recursive subgoals— e.g. Empty Truck reduced to get a package out, Empty Truck, conditions may not have any real connection to abstraction
This is about the only thingThat we can say about HTNsIf we don’t put any restrictionsOn the schemas. However, the question of Validation of schemas then isExternal to the planner since Schemas are supposedly Capturing what kinds of solutionsUsers like. It is not very clear that HTNs are the best way to express preferences on solutions anyway
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
A concretization is a task network produced by repeatedly reducing a non-primitive task until all tasks are primitive
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Modeling Action Reduction
GobyBus(Phx,Msn)Get(Money) Buy(WiscCheese)
At(Msn)
Hv-Money
t1: Getin(B,Phx)
t2: BuyTickt(B)
t3: Getout(B,Msn)
In(B)Hv-Tkt
Hv-MoneyAt(D)
Get(Money)
Buy(WiscCheese)
GobyBus(S,D)
t1: Getin(B,S)
t2: BuyTickt(B)
t3: Getout(B,D)
In(B)
Hv-Tkt
Hv-Money At(D)
Affi
nity
bet
wee
n re
duct
ion
sche
mas
and
plan
-spa
ce p
lann
ing
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Dual views of HTN planning
Capturing hierarchical structure of the domain
– Motivates top-down planning» Start with abstract plans,
and reduce them Many technical headaches
– Respecting user-intent, maintaining systematicity and minimality
[Kambhampati et. al. AAAI-98]
» Phantomization, filters, promiscuity, downward-unlinearizability..
Capturing expert advice about desirable solutions
– Motivates bottom-up planning
» Ensure that each partial plan being considered is “legal” with respect to the reduction schemas
» Directly usable with disjunctive planning approaches
Connection to efficiency is not obvious
Relative advantages are still unclear... [Barrett, 97]
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
2/12
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Changes to Plan-Space Refinement(in the presence of non-primitive tasks)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
HTN vs. T-STN
Nau’s book talks about a “special” version of HTN called T-STN (Total-ordering STN)
T-STN schemas are HTN schemas where all the ordering relations between tasks are constrained to be “contiguity” constraints
– This means that the concretizations of different tasks cannot be interleaved– Bug or feature?
» From the planner’s point of view, it can be a feature Most of the difficulties in HTN planning come from the need to do plan-space refinement in the presence of non-primitive tasks (phantomization, conflict resolution etc..); eliminating that will make HTN planning a no-brainer
» BUT from the domain-writer’s point of view it is clearly a BUG Would have to write many more schemas
– Same is true of Yang’s unique main sub-action thing (although it is less drastic than T-STN)
(few cooking recipes are given with contiguity orderings—which means that you can’t do anything else while following the recipe)
The only HTN planners that use STN schemas are those from Nau’s group …
– SIPE and O-Plan use schemas with precedence orderings
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
HTN vs. STRIPS planning:Competing or Complementary?
Much of the discussion till now has been done with the assumption that HTN is just a way of doing STRIPS planning
HTN can also be seen as solving a different problem– Can develop plans at
various levels of abstraction
– Can model domains with partial domain knowledge
» What you consider “primitive” action may be changeable
What you thought was a correct plan may not be..
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
9/6
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Full procedural control: The SHOP way
Travel by bus only if going by taxi doesn’t work outCan’t do any other actions between !hail and !ride
Shop provides a “high-level” programminglanguage in which the user can code his/herdomain specific planner
-- Similarities to HTN planning Uses T-STN schemas -- Order of use of reduction schemas can be specified The SHOP engine can beseen as an interpreterfor this language
[Nau et. al., 99]
Blurs the domain-specific/domain-independent divideHow often does one have this level of knowledge about a domain?
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Non-HTN Declarative Guidance
Invariants: A truck is at only one location at(truck, loc1, I) & loc1 != loc2 => ~at(truck, loc2, I)
Optimality: Do not return a package to a location at(pkg, loc, I) & ~at(pkg,loc,I+1) & I<J => ~at(pkg,loc,j)
Simplifying: Once a truck is loaded, it should immediately move ~in(pkg,truck,I) & in(pkg,truck,I+1) & at(truck, loc, I+1) => ~at(truck, loc, I+2)
Once again, additional clauses first increase the encoding size but make them easier to solve after simplification (unit-propagation etc).
[Kautz & Selman, AIPS-98]
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Case Study: SATPLAN with domain specific knowledge
Instantiate the domain specific rules as well, and then add them to the encoding.
Solve the full encoding…– Won’t the encoding size increase with domain spefic rules?
» Yes, but the new rules support a lot of simplification so that after simplification, the encoding size actually reduces!!
Example in Blocks world (from Kautz & Selman, AIPS-99)
BW-c:
Original: 4258 vars (10% increase) / 30986 clauses (300% increase)
Final (after simplification): 3526 vars 22535 clauses
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
SAT encodings of HTN planning Abstract actions can be seen as disjunctive constraints
– K-step encoding has each of the steps mapped to a disjunction of the non-primitive tasks
– If a step s is mapped to a task N, then one of the reductions of N must hold (**The heart of encoding setup**)
– + The normal constraints of primitive action-based encoding» Causal encodings seem to be a natural fit (given the causal
dependencies encoded in reduction schemas)
[Mali & Kambampati, AIPS-98]
Constraintsfrom action-based
encodings
HTN constraints
HTN
-com
plian
tS
olu
tion
s
Solu
tion
s f
or
acti
on
based
en
cod
ing
s
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Solving HTN Encodings
[Kambhampati & Mali, AIPS-98]
# ClausesTIME
Puzzle: How can increasing encoding sizes lead to efficient planning?Abstract actions and their reductions put restrictions on the amount of step-action disjunction at the primitive level. --Reduction in step-action disjunction propagates e.g. Fewer causal-link variables, Fewer exclusion clauses…
Savings won’t hold if each non-primitive task has MANY reductions
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Rules on desirable State Sequences: TLPlan approach
TLPlan [Bacchus & Kabanza, 95/98] controls a forward state-space planner
Rules are written on state sequences using the linear temporal logic (LTL)
LTL is an extension of prop logic with temporal modalities U until [] always O next <> eventually
Example:
If you achieve on(B,A), then preserve it until On(C,B) is achieved:
[] ( on(B,A) => on(B,A) U on(C,B) )
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Keep growing “good” towers, and avoid “bad” towers
Good towers are those that do not violate any goal conditions
TLPLAN Rules can get quite boroque
How “Obvious”
are these ru
les?
The heart of TLPlan is the ability to incrementally and effectively evaluate the truth of LTL formulas.
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Control rules for choice points
UCPOP developed a language in which user can write control rules that tell the planner what to do when it faces a specific sort of choice (e.g. between open conditions; establishers for open conditions etc.).
HSTS—a metric temporal planner that NASA used in its DeepSpace mission used similar control rule language.– The user needs to know the innards of the planner to write
these rules– The rules are hard to debug
» Learning such rules automatically is a way out
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Folding the Control Knowledge into the planner: CLAY approach
Control knowledge similar to TLPlan’s
Knowledge is folded using KIDS semi-automated software synthesis tool into a generic refinement planning template– Use of program optimizations
such as » Finite differencing» Context-dependent &
independent simplification Empirical results demonstrate that
folding can be better than interpreting rules
Srivastava & Kambhampati, JAIR 97
Caveat: Current automated software
synthesis tools
have a very steep learning curve
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Many User-Customizable Planners Conjunctive planners
– HTN planners » SIPE [Wilkins, 85-]» NONLIN/O-Plan [Tate et. al., 77-]» NOAH [Sacerdoti, 75]» Also SHOP (Nau et. al., IJCAI-99)
– State-space planners» TLPlan [Bacchus & Kabanza, 95; 99]» TALPlan [Kvarnstrom & Doherty, 2000]
– Customization frameworks
» CLAY [Srivastava & Kambhampati, 97] Disjunctive planners
– HTN SAT [Mali & Kambhampati, 98] – SATPLAN+Dom [Kautz & Selman, 98]
How do th
ey rela
te?
What a
re th
e tradeoffs?
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
With right domain knowledge any level of performance can be achieved...
HTN-SAT, SATPLAN+DOM beat SATPLAN…– Expect reduction schemas, declarative
knowledge about inoptimal plans TLPLAN beats SATPLAN, GRAPHPLAN
– But uses quite detailed domain knowledge
SHOP beats TLPLAN…(but not TALPlan)– Expects user to write a “program” for the
domain in its language » Explicit instructions on the order in
which schemas are considered and concatenated
Who g
ets
the c
redit
?
-The p
rovi
der
of dom
ain
know
ledge?
-The p
lanner
that
is c
apable
of usi
ng it?
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Types of Guidance
Declarative knowledge about desirable or undesirable solutions and partial solutions (SATPLAN+DOM; Cutting Planes)
Declarative knowledge about desirable/undesirable search paths (TLPlan & TALPlan)
A declarative grammar of desirable solutions (HTN)
Procedural knowledge about how the search for the solution should be organized (SHOP)
Search control rules for guiding choice points in the planner’s search (NASA RAX)
Planner specific. Expert needs to understand the specific details of the planner’s search space
(largely) independent of the details of the specific planner[affinities do exist between specific types of guidance and planners)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Ways of using the Domain Knowledge As search control
– HTN schemas, TLPlan rules, SHOP procedures– Issues of Efficient Matching
To prune unpromising partial solutions– HTN schemas, TLPlan rules, SHOP procedures – Issues of maintaining multiple parses
As declartative axioms that are used along with other knowledge– SATPlan+Domain specific knowledge– Cutting Planes (for ILP encodings) – Issues of domain-knowledge driven simplification
Folded into the domain-independent algorithm to generate a new domain-customized planner– CLAY – Issues of Program synthesis
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Conundrums of user-assisted cutomization
Which planners are easier to control?– Conjunctive planners are better if you have search control knowledge
» Forward State Space (according to TLPlan)
» Plan-space planners (according to HTN approaches)
– Disjunctive planners are better if your knowledge can be posed as additional constraints on the valid plans
» Which SAT encoding? HTN knowledge is easier to add on top of causal encodings
Which approach provides the best language for expressing domain knowledge for the lay user?– (Mine--no, no, Mine!)
What type of domain knowledge is easier to validate? When does it become “cheating”/ “wishful-thinking”
– Foolish not to be able to use available knowledge
– Wishful to expect deep procedural knowledge...
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
“Evaluating” KB-planners
Since the 2nd IPC, there has been a separate track for “knowledge-based planners” at the competition.– Basically only the TLPlan (and its variant TALPlan) and SHOP took
part (in comparison close to a dozen planners take part in the domain-independent track)
» In 2000, TALPlan won» In 2002, TLPlan won
– Quite controversial When SHOP does better than TLplan on Gizmo domain, what
does this mean?– That SHOP is a better planning algorithm than TLplan?– That Dana Nau knows Gizmo domain better than Faheim Bacchus?– That the easily available control knowledge in Gizmo domain is
more naturally representable as TLplan rules than SHOP schemas?
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Are we comparing Dana & Fahiem or SHOP and TLPlan?
(A Critique of Knowledge-based Planning Track at ICP)
Subbarao KambhampatiDept. of Computer Science & Engg.
Arizona State UniversityTempe AZ 85287-5406
Click here to download TLPlan– Click here to download a
Fahiem
Click here to download SHOP– Click here to download a
Dana
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
The “I am not an anti-dentite”Disclaimers..
I think KB planning is a swell idea– I started my career with HTN planning…
I think the KB planning track at IPC is a swell idea– has done more to increase interest in KBplanning than the bi-annual polemics and laments about lack of
interest in “Knowledge-based planning”
I think Fahiem and Dana are REALLY swell – (in case they don’t buy that) I may already have a black-belt in Karate..
I’d rather learn from one bird how to singthan teach ten thousand stars how not to dance
--e.e. cummings
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
What are the lessons of KB Track?
If TLPlan did better than SHOP in ICP, then how are we supposed to interpret it?– That TLPlan is a superior planning technology over SHOP? – That the naturally available domain knowledge in the
competition domains is easier to encode as linear temporal logic statements on state sequences than as procedures in the SHOP language?
– That Fahiem Bacchus and Jonas Kvarnstrom are way better at coming up with domain knowledge for blocks world (and other competition domains) than Dana Nau?
We are NOT asking the right questions
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Questions worth asking in KB planner comparisons (IMHO)
How easy/natural (for humans) is the language in which the planner accepts control knowledge?
How easy is it to “validate” the control knowledge being input to the planner?
Is the naturally available knowledge about a specific domain easily encoded in the language accepted by the planner?
Does the planner allow “any expertise” behavior—solving the problems even without any control knowledge, but improving performance with added control knowledge? (or is the control knowledge tightly intertwined with the domain physics?).
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
How/Why the competition is not asking the right questions…
The role of the knowledge-engineer is played by the same person(s) who wrote the planner. So, the question of how natural the specific language is for third-party knowledge engineers is largely unaddressed.
No reasonable time limits are placed on coming up with the control knowledge. So, we don’t learn much (or anything) about whether or not naturally available knowledge about a domain is easily representable in the language accepted by the planner.
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Some suggestions for change…
Recruit third-party volunteers who will play the role of knowledge engineers for the KB planners.
– Ideally, we would like to have the same people writing the control knowledge for a given domain for all the competing approaches (so one knowledge engineer per domain rather than one knowledge engineer per planner).
(Alternative to above) Specify the control knowledge that is available, so all planners encode the same general knowledge.
– One idea might be to ask the designers of the domains (e.g. David Smith and his cohorts for the Satellite and Rovers domain) to provide, in english, what sort of control information they would like the planner to use.
Measure the time taken to write and validate the control knowledge. Analyze the knowledge encoded by the different KB planners for the
same domain– Characterize it in terms of (a) whether the knowledge is procedural or
declarative and (b) how hard would it be to “learn” the same knowledge.
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Slides from here-on about “learning” domain knowledge were not used
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Automated Customization of Planners
Domain pre-processing– Invariant detection; Relevance detection;
Choice elimination, Type analysis
» STAN/TIM, DISCOPLAN etc.
» RIFO; ONLP Abstraction
» ALPINE; ABSTRIPS, STAN/TIM etc. Learning Search Control rules
» UCPOP+EBL,
» PRODIGY+EBL, (Graphplan+EBL) Case-based planning (plan reuse)
» DerSNLP, Prodigy/Analogy
Mos
t of
the
wor
k ha
s be
en d
one
in the
cont
ext of
Con
junc
tive
Refi
nem
ent Pl
anne
rs
Read Terry Zimmerman’s “Learning assisted planning: Looking back, Taking Stock and Going Forward”
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Symmetry & Invariant Detection
Generate potential invariants and test them – DISCOPLAN [Gerevini et. al.]
» Allows detection of higher-order mutexes
– Rintanen’s planner
» Uses model-verification techniques
– STAN/TIM
» Type analysis of the domain is used to generate invariants
– ONLP (Peot & Smith)
» Use operator graph analysis to eliminate non-viable choices
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Abstraction
Idea– Abstract some details of the problem or actions. – Solve the abstracted version. – Extend the solution to the detailed version
Precondition Abstraction– Work on satisfying important preconditions first
» Importance judged by: Length of plans for subgoals [ABSTRIPS, PABLO] Inter-goal relations [ALPINE] Distribution-based [HighPoint]
– Strong abstractions (with downward refinement property) are rare– Effectiveness is planner-dependent
» Clashes with other heuristics such as “most constrained first”
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Example: Abstracting Resources
Most planners thrash by addressing planning and scheduling considerations together
– Eg. Blocks world, with multiple robot hands
Idea: Abstract resources away during planning
– Plan assuming infinite resources
– Do a post-planning resource allocation phase
– Re-plan if needed
(with Biplav Srivastava)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Learning Search Control Rules with EBL
UCPOP+EBL
Explain leaf level failures Regress the explanationsto compute interior nodefailure explanations
Use failure explanations toset up control rules Problems: -- Most branches end in depth-limits >No analytical explanation >Use preference rules? -- THE Utility problem >Learn general rules >Keep usage statistics & prune useless rules
If Polished(x)@S & ~Initially-True(Polished(x)) Then REJECT Stepadd(Roll(x),Cylindrical(x)@s)
(Kambhampati, Katukam, Qu, 95)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Case-based Planning Macrops, Reuse, Replay
Structures being reused– Opaque vs. Modifiable
– Solution vs. Solving process (derivation) Acquisition of structures to be reused
– Human given vs. Automatically acquired Mechanics of reuse
– Phased vs. simultaneous Costs
– Storage & Retrieval costs; Solution quality
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Case-study: DerSNLP
Modifiable derivational traces are reused Traces are automatically acquired during problem solving
– Analyze the interactions among the parts of a plan, and store plans for non-interacting subgoals separately
» Reduces retrieval cost– Use of EBL failure analysis to detect interactions
All relevant trace fragments are retrieved and replayed before the control is given to from-scratch planner
– Extension failures are traced to individual replayed traces, and their storage indices are modified appropriately
» Improves retrieval accuracy
( Ihrig & Kambhampati, JAIR 97)
Oldcases
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
DerSNLP: ResultsC
um
ula
tive
CP
U T
ime
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
1 2 3 4 5 6
0
20
40
60120
% S
olv
ed
01020
30405060
708090
100
0
20 40 60 80
100
120
Lib
rary
Siz
e
0100200300400500600
0
20
40
60
80
100
120
Performance withincreased Training
% Solvability withincreased traning
(JAIR, 97)
Lib
rary
Siz
e
5 31
Problem size (#goals)
MA2: Recent Advances in AI Planning: A Unified View Subbarao Kambhampati
Reuse in Disjunctive Planning Harder to make a disjunctive planner commit to
extending a specific plan first Options:
– Support opaque macros along with primitive actions» Increases the size of k-step disjunctive plan» But a solution may be found at smaller k
– Modify the problem/domain specification so the old plan’s constraints will be respected in any solution (Bailotti et. al.)
– MAX-SAT formulations of reuse problem» Constrain the encoding so that certain steps may have
smaller step-action mapping and ordering choices» Causal encodings provide better support
[with Amol Mali]